Actuarial Questions

Title page
Australian School of Business

Probability and Statistics
Tutorial Exercises

Exercises
Session 1, 2014
Overview

Overview
Tutorials
Students must attend the tutorial for which they are enrolled. Attendance will be recorded and count towards meeting the requirements to pass the course. If you wish to change your tutorial then you must lodge
an application to change your tutorial time with the ASB student service centre.
In tutorials, we will implement interactive learning where collaborative group work is highly encouraged.
To get the most out of the tutorials, students should read lecture notes and textbooks and references and
complete assigned homework problems in advance of the tutorial.
Peer-Assisted-Support-Scheme (PASS)
In addition, ASOC has an actuarial PASS program will be available to enhance student learning in this
course. It is highly recommended that you attend these sessions.
ASOC website: www.asoc.unsw.edu.au
PASS peer support class times:
www.asoc.unsw.edu.au/index.php?option=com content&view=article&id=48&Itemid=42
PASS material is available on: sites.google.com/a/asoc.unsw.edu.au/unsw-asoc/downloads
Organization of Tutorials
The more you read the more you know, but the more you practice the more you learn and understand. So
the key to the understanding of this course is problem solving.
The purpose of tutorials is to enable you to raise questions about difficult topics or problems encountered
in their studies. Students must not expect another lecture, tutorials are a place where you have to learn
by doing and subsequently also learn from mistakes, as that is the most efficient way to learn, especially in
actuarial courses.
To benefit from the tutorial exercises provided, you have to make the attempt to the questions yourself.
Copying the tutorial manuals solution or the tutors solution is least effective for your learning process.
Moreover, given that you already know the answer, the copying learning strategy prevent you from acquiring the knowledge of solving the question when trying to solve the exercise any subsequent time afterwards.
The object of the tutorials is to attempt the assigned exercises and topics covered in the course. Each
week a document will be posted containing the exercises which are to be covered in tutorials. IN the schedule
on the next page the weekly questions are divided into three groups:
1. Exercises that should be done in the same week as the lecture. At the end of the week the solutions
to these questions will be uploaded on Blackboard. If there is enough interest there will be an online
tutorial about those questions. Link and details will be provided via Blackboard.
2. Question done in the week after the lecture. The class tutorial will typically focus on these questions.
After all tutorials, a new solution manual will be posted on Blackboard containing all the answers.
3. Some optional/additinal/exam preparation questions for the ones requiring more exercises than only
the tutorial, pass class and book exercises when studying for the exam.
Tutorials are an integral component of the course; attendance and participation in your tutorial is crucial
for your successful completion of the course.
The Teaching Strategy (including the feedback loop) in this course is:
The Get introduced (explanation of course concepts in the lecture) and Try it out (examples in the
lecture) are part of the lecture. If you are not able to make satisfactory attempt to the examples in the
lecture, this is feedback that you should revise the lecture material in dept after the lecture. The Try again
(tutorial exercises) and Get feedback (answers from tutor/ Wikispace / tutorial solutions) are part of the
tutorial. The tutorial is designed to attempt the tutorial questions in groups (typically of 3-4 students). This
collaborative working is advised since it allows learning from peers and allows the more advanced student
with a possibility to test whether s/he knows the material in dept and is thereby able to explain it to peers.
c Katja Ignatieva
!
School of Risk and Actuarial Studies, ASB, UNSW
Page i of iii
ACTL2131 & ACTL5101
Get
Introduced
Try it out
Get feedback
Overview
Try again
Figure 1: Feedback loop
If the group is unable to solve a question, it can ask for help from the tutor. The tutor will not provide the
answer, but would help you in the direction of the solution. This is because you should practice and learn
by doing, rather than seeing the solution. At the end of the week the tutorial solutions will be posted on
Blackboard. It will only be posted at the end of the week to give you time to attempt the questions without
a solution manual. The solution manual should not be part of attempting a question, but to verify whether
your attempt was correct.
Learning Strategy
A required learning strategy for the tutorials (on which provision of the course materials is based) is:
1) Prior to make an attempt of the exercises, review your lecture notes.
2) Prior to the tutorial, make an attempt to the exercises you should do before the tutorial (see schedule
below).
3) At the end of the week, verify your answers with the (partial) solution manual and/or the online
tutorial on Blackboard. Revise the material that you did not know, did incorrectly.
4) During the tutorial, make an attempt to the exercises you should do in the tutorial (see schedule
below).
5) After the tutorial, make an attempt to the exercises you should do after in the tutorial (see schedule
below).
6) If you have questions about the tutorial exercises, ask them at the Wikispace. If you think you have
a good understanding of the material, you should try and answer the questions of your peers on the
Wikispace. This will give you feedback on your ability to explain the material and hence how well you
know the material.
7) Check your answer using the tutorial solution manual.
NOTE: tutorials are for making tutorial exercises, not for reviewing material or working on
assignment or anything else!
Schedule of Tutorial Exercises
c Katja Ignatieva
!
Page ii of iii
ACTL2131 & ACTL5101
Exercises
Week
Week
Week
Week
Week
1
2
3
4
5
Week 6
Week 7
Week 8
Week 9
Week 10
Week 11
Week 12
Sample Exam
Overview
Before tutorial
1, 2
1, 2
1, 2, 3
1
1, 2
During Tutorial
After Tutorial
Additional
10, 11, 12, 13

3, 4, 5
4, 6, 7, 5
3abc, 2, 4, 3d,e
3, 4, 5, 6
14, 15, 16, 17

6, 7, 8
8, 9, 10, 11
5, 6
7, 8, 9
3-9
1, 2
1
Working on
assignment
Working on
assignment
Working on
assignment
Working on
assignment
15 of week 11
tutorial
Revision for
exam
4, 3, 6, 5
2, 3, 4, 6
1, 2, 8, 7
7, 8, 9, 10
5, 7
Working on
ment
Working on
ment
Working on
ment
Working on
ment
Revision for
1, 2a, 5
2-6
5-14
1-3
12
10, 11, 12, 13, 14,
15, 16
11, 12, 13, 14
assign-
3, 5, 6
assign-
3a-d, 4, 7a-d
assignassign-
1, 7, 8
exam
1-4
Revision for exam
Table 1: Schedule of tutorial exercises.
c Katja Ignatieva
!
Page iii of iii
Week 1

Tutorial Exercises, Week 1, 2014
1. An urn contains one black ball and one gold ball while a second urn contains one white and one gold
ball. One ball is selected at random from each urn.
(a) Describe the sample space for this experiment.
(b) Describe a -algebra for this experiment.
(c) Describe the event that both balls will be of the same colour. What is the probability of this
event?
2. A box contains 100 Christmas balls: 49 are red, 34 are gold, and 17 are silver. Three balls are to be
drawn without replacement. Determine the probability that:
(a) all 3 balls are red;
(b) the balls are drawn in the order: red, gold, and silver;
(c) the third ball is a silver, given that the first 2 are red and gold (not necessarily in that order);
and
(d) the first 2 are red, given that the third ball is a silver;
3. Let A and B be two independent events. Prove that the following pairs are also independent:
(a) A and B C
(b) AC and B
(c) AC and B C
4. A pair of events A and B cannot be simultaneously mutually exclusive and independent. Assume that
their probabilities are strictly positive, i.e., Pr (A) > 0 and Pr (B) > 0. Prove the following:
(a) If A and B are mutually exclusive, then they cannot be independent.
(b) If A and B are independent, then they cannot be mutually exclusive.
5. This exercise shows that independence does not imply pairwise independence. Consider a random
experiment which consists of tossing two dice. Define the following events:
E1
E2
E3
(a)
(b)
(c)
(d)
= {doubles appear}
= {the sum is between (and includes) 7 and 10}

= {the sum is 2 or 7 or 8}
Show that E1 , E2 and E3 are independent.

Show that E1 and E2 are not pairwise independent.
Show that E2 and E3 are not pairwise independent.
What about E1 and E3 are they pairwise independent?
6. In an undergraduate statistics class, three students A, B, and C submitted exactly (word-for-word)

the same solution to a homework problem. It is the policy of the lecturer to give zero marks for those
who copy homework problems. Believing that there must be one of the three who actually did the
work, the lecturer will pardon one of the three and chooses at random the student to pardon.
However, the lecturer will only inform the students at the end of the semester who among them has
been pardoned.
The next day, A tries to get the lecturer to tell him who had been pardoned. The lecturer refuses. A
then asks which of B or C will not be pardoned. The lecturer thinks for a while, then tells A that B
is not to be pardoned.
c Katja Ignatieva
!
Page 1 of 4
ACTL2131 & ACTL5101
Tutorial Exercises, Week 1
Lecturers reasoning: Each student has a 1/3 chance of being pardoned. Clearly, either B or C
must not be pardoned, so I have given A no information about whether A will be pardoned.
As reasoning: Given that B will not be pardoned, then either A or C will be pardoned. My
chance of being pardoned has risen to 1/2.
(a) Evaluate the lecturers reasoning, i.e., explain whether his reasoning is justified.
(b) Evaluate student As reasoning, i.e., explain whether his reasoning is justified.
7. Two airlines serving some of the same cities in Australia have merged. Management has decided to
eliminate some of the repetitious daily flights. On the Perth-Sydney route, one airline originally had
five daily flights (each at different a time) and the other had six daily flights (each at different a time).
Determine the number of ways:
(a) four flights can be eliminated.
(b) the first airline can eliminate two of its scheduled five flights.
(c) the second airline can eliminate two of its scheduled six flights.
(d) two flights can be eliminated from each airline.
8. Three boxes are numbered 1, 2 and 3. For k = 1, 2 and 3, box k contains k blue marbles and 5 k
red marbles. In a two-step experiment, a box is selected and 2 marbles are drawn from it without
replacement. If the probability of selecting box k is proportional to k, what is the probability that the
two marbles drawn have different colors?
9. The probability function of a certain discrete random variable on the non-negative integers satisfies
the following:
Pr(0) = Pr(1)
Pr(k + 1) = Pr(k)/k
for k = 1, 2, 3, . . ..
Find Pr(0).
10. Consider X, a continuous random variable with density function
fX (x) = cex ,
x > 1, and zero otherwise.
Find
(a) all c such that fx is a random variable, and
(b) Pr(X < 3 | X > 2).
11. The distribution function for a discrete random variable X is given by:
if x < 1
0
1/3
if 1 x < 2/3
FX (x) =
1
if x 2/3.
(a) Specify the probability mass function pX (x).
(b) Sketch the graphs of pX (x) and FX (x).

12. Let X be a random variable with density:
$
%
&2 '
1 x
1
exp
fX (x) =
2
, for < x < .
Here, X is called a normally distributed random variable.

(a) Find an expression for the moment generating function, MX (t) of X.
(b) Now define S (t) = log [MX (t)]. Show that, in general,
(
(
(
(
d2
d
(
S (t)((
S
(t)
= E [X] and
= V ar (X) .
(
2
dt
dt
t=0
t=0
c Katja Ignatieva
!
Page 2 of 4
ACTL2131 & ACTL5101
(c) Use the above result to prove that, with the normal density, we have
E (X) =
and V ar (X) = 2 .
(d) How do we call the function S (t)?

13. Let X be a random variable with parameters , , , and (, and have the following moment
generating function: MX (t) = t + t2 + t3 + t4 .
(a)
(b)
(c)
(d)
(e)
How many distribution functions corresponds to this m.g.f. for given values of the parameters?
Determine the first five non-central moments of X.
Determine the first five central moments of X.
Determine the mean, variance, skewness, and kurtosis of X.
Let X represents the claim sizes, i.e., a higher value is bad for the insurer. Insurer A and Bi
ask a quote for reinsuring a tail risk (for example: the reinsurer makes a payment to the insurer
if the loss is larger than $1 million). Based on the mean, variance, skewness and kurtosis, which
of the two would receive a higher quote for reinsuring the risk, and why, if:
i) As parameters are: = 1, = 2, = 1, and = 1 and B1 parameters are = 1,
= 1, = 0.5384, and = 0.2606;
ii) As parameters are: = 1, = 2, = 1, and = 1 and B2 parameters are = 1, = 2,
= 2, and = 2;
iii) As parameters are: = 1, = 2, = 1, and = 1 and B2 parameters are = 1, = 2,
= 1, and = 1.625.
14. The probability density function for a continuous random variable X is given by:
)
2/x3
for x 1
fX (x) =
0
otherwise.
(a) Determine a formula for the cumulative distribution function FX (x).
(b) Determine the probability that X 4.
(c) Sketch the graphs of fX (x) and FX (x).
15. Let X be a random variable with probability density function:
1 x
, if x 0;
2 e
fX (x) =
1 x
if x < 0.
2 e ,
(a)
(b)
(c)
(d)
Verify that fX () is a pdf.

Find expression for the cdf FX (x).
Find the moment generating function and the probability generating function of X.
Suppose = 1. Evaluate Pr (|X| < 3/4).
16. Actuaries often model the age-at-death as a non-negative random variable X and define the force of
mortality as follows:
(x) = lim
h0
FX (x + h) FX (x)
h (1 FX (x))
where FX () denotes the cdf of X.

(a) Using this definition, prove that:
% *
FX (x) = 1 exp
&
(z) dz .
(b) Show that for a non-negative random variable:

*
E [X] =
[1 FX (z)] dz.
0
Use this result to show that:

E [X] = E
c Katja Ignatieva
!
,
1
.
(X)
Page 3 of 4
ACTL2131 & ACTL5101
17. A random variable X has a probability density function of the form:

.
fX (x) = ax 1 bx2 , for 0 x 1, and zero otherwise,
where a and b are positive constants.
(a) Determine the value of a in terms of b and show that b 1.
(b) For the case b = 1, determine the mean and variance of X.
-End of week 1 Tutorial Exercises-
c Katja Ignatieva
!
Page 4 of 4
Week 2

1. For each of the following situations, specify the type of distribution that best models the random
variable X and give the parameters of the distribution chosen (where possible):
(a) This year, there are 100 students enrolled in an introductory actuarial studies course. For the
mid-session test for this course, the papers are marked by a team of tutors; however, a sample
of these papers is examined by the course professor for marking consistency. Experience suggests
that 1% of all papers will be improperly marked. The professor selects 10 papers at random from
the 100 papers and examines them for marking inconsistencies. X is the number of papers in the
sample that are improperly marked.
(b) A standard drug has been known to be effective in 90% of the cases in which it is used. To
re-evaluate the effectiveness of this same drug, a clinical trial will be performed where 20 has
volunteered. X is the number of cases where the drug has been found effective.
(c) An immunologist is studying blood disorders exhibited by people with rare blood types. It is
estimated that 10% of the population has the type of blood being investigated. Volunteers whose
blood type is unknown are tested until 100 people with the desired blood type are found. X is
the number of people tested who do not have the desired rare blood type.
(d) Customers arrive at a fastfood restaurant independently and at random. During lunch hour,
where more customers are often expected to arrive, customers arrive at the fastfood restaurant
at the rate of two per minute on the average. X is the number of people who arrive between 12:15
p.m. and 12:30 p.m.
(e) A set of 25 multiple-choice questions was asked in an examination. It has been determined,
according to experience, that the proportion of the questions which are guessed and answered
correctly is 35%. X is the number of questions guessed and answered correctly by a particular
student who wrote for the examination.
2. For each of the following moment generating functions of discrete random variables X, identify the
distribution and specify the associated parameters.
et
2 et
% t
&3
e +1
MX (t) =
2
%
&
1 t 1
MX (t) = exp
e
2
2
% t &4
e
MX (t) =
2 et
% t
&5
3e + 1
MX (t) =
4
(a) MX (t) =
(b)
(c)
(d)
(e)
3. Poisson approximation to the binomial. This exercise is to show that binomial probabilities can
be approximated using the Poisson probabilities, which are generally easier to calculate. Let X
Binomial(n, p) and Y Poisson() where = np. The approximation states that
Pr (X = x) Pr (Y = x) ,
for large n and small np. This can be proven using convergence of mgfs. Denote the respective mgfs
by MX (t) and MY (t).
c Katja Ignatieva
!
Page 21 of 23
ACTL2131 & ACTL5101
(a) Prove that limn MX (t) = MY (t).

1
Hint: use lim (1 + nx )n = exp(x) = lim (1 + hx) h .
n
h0
(b) Another method to prove this approximation is as follows: First, establish that the Poisson
distribution satisfies the relation
Pr (Y = x)
=
Pr (Y = x 1)
x
for x = 1, 2, . . . .
Second, a similar relation can be approximated for the binomial distribution:

lim
p0
Hint: first show that
Pr(X=x)
Pr(X=x1)
Pr (X = x)
np
=
.
Pr (X = x 1)
x
nx+1
x
show that:
p
1p ,
then take lim and use that lim (np) = . Then

p0
p0
Pr (Y = 0) Pr (X = 0) ,
for large n.
(c) A typesetter, on the average makes one error in every 400 words typeset. A typical page contains
300 words. Use the Poisson approximation to the binomial to compute the probability that there
will be more than 3 errors in 10 pages.
4. An insurance company receives 200 claims per day on the average. Claims arrive independently and
at random at the company office. Of the claims, 95% are for amounts less than $100 and are processed
immediately; the remaining 5% are examined more closely to verify their accuracy and eligibility.
(a) Determine the probability of getting no claims over $100 in a given day.
(b) Determine the probability of getting at most two claims over $100 in a given day.
(c) How many claims for amounts less than $100 should this company expect to receive in a week (5
business days)?
5. Let X have a Gamma(, ) distribution.
(a) Prove that the mgf of X can be expressed as:
%
&
for t < .
(b) Establish also that for any positive constant r
E [X r ] = r
(r + )
.
()
6. Suppose that you have $1 000 to invest for a year. You are currently evaluating two investments:
Investment A and Investment B, with annual rates of return, respectively denoted by RA and RB .
Assume:
RA Normal (0.05, 0.1) and RB Normal (0.10, 0.5) .
(a) Under Investment A, compute the probability that your investment will be below $1 000 in a year.
(b) Under Investment A, compute the probability that your investment will exceed $1 200 in a year.
(c) Under Investment B, compute the probability that your investment will below $1 000 in a year.
(d) Under Investment B, compute the probability that your investment will exceed $1 200 in a year.
7. A city engineer has studied the frequency of accidents at two busy intersections. He has determined
that the time T in months between accidents at each intersection has an exponential distribution. The
parameters for these two distributions are 2 and 2.5. Assume that the occurrence of accidents at these
intersections is independent.
(a) Determine the probability that there are no accidents at either intersection in the next month.
c Katja Ignatieva
!
Page 22 of 23
ACTL2131 & ACTL5101
(b) Determine the probability that there will be no accidents for at least one of the intersections in
the next month.
8. The Pareto distribution is very commonly used to model certain insurance loss amounts. We say X
has a Pareto distribution if its density can be expressed as:
fX (x) =
% &+1
for x > ,
x
and zero otherwise.

(a) Find expressions for the mean and variance of X.
1
(b) Find expression
the. quantile of X. The quantile function is f (u) = FX
(u) hence, one should
- for
1
solve u = FX FX (u) .
(c) What is then its median (i.e., u = 0.5)?
(d) An insurance policy has a deductible1 of $5. The random variable for the loss amount (before
deductible) on claims filed has a Pareto distribution with = 3.5 and = 4. Find:
1. the mean loss amount;
2. the expected value of the amount of a single claim; and
3. the variance of the amount of a single claim.
1 A deductible is that the policy only makes no payment if the loss amount is smaller than the deductible; and the claim
amount equals the loss amount minus the deductible if the loss amount is larger than the deductible.
c Katja Ignatieva
!
Page 23 of 23
Week 3

1. The claim amounts (in dollars, to the nearest $10) for a sample of 24 recent claims for storm damage
to private homes in a particular town are as follows:2
2 710
670 2 380 4 670 1 220 6 780 1 590 3 110
960 8 230 3 320 3 380 2 490 1 940 3 710 4 630
4 270 4 210 1 880 3 880 1 490 5 400 2 430
850
(a) Construct a stem-and-leaf display of these claim amounts.
(b) Find the mean and median of the claim amounts. What can you say about the skewness of the
distribution?
(c) Find the interquartile range of the claim amounts.
(d) Evaluate F24 (1 000) where F24 () denotes the ecdf.
2. Data were collected on 100 consecutive days for the number of claims, X, arising from a group of
insurance policies. This resulted in the following frequency distribution:
observed claims from policy (x):
frequency:
0
14
1
25
2
26
3
18
4
12
5
5
Calculate the following sample statistics for these data:

(a) mode
(b) median
(c) interquartile range
(d) Suppose the average value for 5 claims or more is 7.5. Calculate the sample mean.
3. For a set of 32 observations, you are given:
32
/
xk = 13 337.6
and
k=1
32
/
x2k = 5 667 388.7.
k=1
The largest of the observations is 605. Suppose you are interested in measuring the impact of the
largest observation on the mean and standard deviation.
(a) Calculate the sample mean and the sample standard deviation.
(b) Calculate the sample mean and the sample standard deviation, with the largest observation
deleted.
(c) What is the percentage change in the mean?
(d) What is the percentage change in the standard deviation?
4. Let X and Y be two discrete random variables whose joint probability function is given by:
Pr(X = x, Y = y)
Y =y
2 Modified
1
2
0
0.05
0.20
X=x
1
2
0.20 0.15
0.15 0.12
3
0.05
0.08
Institute of Actuaries exam question.
c Katja Ignatieva
!
Page 31 of 33
ACTL2131 & ACTL5101
Calculate:
(a) E [X]
(b) E [Y ]
(c) E [X |Y = 1 ]
(d) V ar (Y |X = 3 )
(e) E [XY ] and Cov(X,Y).
5. Let X and Y be two discrete random variables whose joint probability mass function is given by:
Pr(X = x, Y = y)
1
2
3
4
Y =y
X=x
2
3
0.05 0.02
0.20 0.05
0.05 0.20
0.02 0.04
1
0.10
0.05
0.02
0.02
4
0.02
0.02
0.04
0.10
(a) Find the marginal probability mass functions of X and Y .

(b) Find the conditional probability mass of X given Y = 2 and of Y given X = 2.
(c) Find E[XY ] and Cov(X, Y ).
6. Let X and Y have the joint density:
fX,Y (x, y) =
6
(x + y)2 , for 0 x 1 and 0 y 1,
7
and zero otherwise.

(a) By integrating over the appropriate regions, find:
1. Pr (X < Y )
2. Pr (X + Y 1)
.
3. Pr X 21
(b) Find the marginal densities of X and Y .

(c) Find the two conditional densities.
7. Two independent measurements, X and Y , are taken of a quantity . We are given the means are
2
equal, E [X] = E [Y ] = , but the variances X
and Y2 are not equal. The two measurements are then
combined by means of a weighted average to give:
Z = X + (1 ) Y,
where is a constant between 0 and 1, i.e., 0 1.
(a) Show that E [Z] = .
(b) Find in terms of X and Y to minimise V ar (Z).
(c) Under what circumstances is it better to use the average (X + Y ) /2 than either X or Y alone to
determine ? Hint: a smaller variance would give a better estimate of the population mean.
(d) Now, suppose X and Y are instead not independent and have covariance:
Cov (X, Y ) = XY .
Find in terms of X , Y and XY to minimise V ar (Z).
8. Let xn and s2n denote the sample mean and variance for the sample x1 , x2 , . . . , xn . Let xn+1 and s2n+1
denote these quantities when an additional observation xn+1 is added to the sample.
(a) Show how xn+1 can be computed from xn and xn+1 .
(b) Show that:
s2n+1 =
n1
n
&
s2n +
1
2
(xn+1 xn )
n+1
so that s2n+1 can be computed from x

n , xn+1 , and s2n .
c Katja Ignatieva
!
Page 32 of 33
ACTL2131 & ACTL5101
9. Suppose X and Y are two continuous random variables. Prove that:

*
E [Y ] =
E [Y |X = x ] fX (x) dx.
10. You are given:

X1 Uniform[0, 1]
Conditional on X1 , X2 Uniform[0, X1 ]
(a) Find the joint distribution function of X1 and X2 .
(b) Find the marginal distribution function of X2 .
11. Suppose that the joint distribution function of X1 and X2 is
0, 1
x1 x2 1 + 21 (1 x1 ) (1 x2 ) ,
FX1 ,X2 (x1 , x2 ) =
Fx1 (x1 ),
Fx2 (x2 ),
given by
if
if
if
if
x1 < 0 or x2 < 0;
0 x1 1 and 0 x2 1;
x2 > 1;
x1 > 1,
(a) Find the joint density.
(b) Find the marginal distribution functions of X1 and X2 . Can you recognise these distributions?
(c) Find the correlation coefficient of X1 and X2 .
12. We have the joint probability density function:
)
k(1 x2 ), if 0 x1 x2 1;
fX1 ,X2 (x1 , x2 ) =
0,
else.
(a) Determine the value k for which this function is a density.
(b) Determine the region for the integral for determining Pr(X1 3/4, X2 1/2).
(c) Calculate Pr(X1 3/4, X2 1/2).
c Katja Ignatieva
!
Page 33 of 33
Week 4

1. Compound Distribution. In a portfolio of insurance policies, the amount of claim is a random variable
1
Xk which has an exponential distribution with mean , for k = 1, 2, . . . The number of claims N in a
single period is also a random variable but with a Poisson() . The total claims then in the portfolio
during the period is given by:
S = X1 + X2 + . . . + XN .
(a) Find the mean of S, E[S].
(b) Find the variance of S, V ar(S).
(c) Find the moment generating function of S, MS (t).
2. Let X1 , X2 and X3 be i.i.d. with common density:
fX (x) = ex ,
x 0.
(a) Find the joint density of X(1) , and X(3) .

Hint: First find the joint density of X(1) , X(2) , and X(3) , i.e., fX(1) ,X(2) ,X(3) (y1 , y2 , y3 ) = . . .
Second you find the distribution of only X(1) and X(3) by integrating over the other random
variable (similar to finding the marginal distribution). Be careful by the limits for X(2) , what are
the lowest and highest numbers it can take?
1
2
1
2
(b) Compute E X(1) and E X(3) .
.
.
(c) Compute V ar X(1) and V ar X(3) .
1
2
.
(d) Compute E X(1) X(3) and the correlation coefficient X(1) , X(3) .
3. Let X Gamma(, 1) and Y Gamma(, 1) be independent random variables. Define U = X + Y

and V = X/(X + Y ).
(a) Use the moment generating function technique to find the distribution of U .
(b) Use the Jacobian transformation technique to find the joint distribution of U and V .
(c) Show that U and V are independent.
Hint: You do not need to do any additional calculations to show this.
(d) Find the marginals of U and V using their joint distribution derived in Question 3b. Demonstrate
the the marginal of U is consistent with that derived from Question 3a.
(e) Use Question 3c. and Question 3d. to find the mean and variance of V .
4. Let X1 , X2 and X3 be three independent and identically distributed as Exp(1) random variables. Find:
(
1
2
(a) E X(3) (X(1) = x
(
1
2
(b) E X(1) (X(3) = x
(c) fX(1) ,X(3) (x, y)
(d) fR (r), where R = X(3) X(1) is the range.
5. Let X1 and X2 be i.i.d. (independent and identically distributed) N (0, 1) random variables.
(a) Show that X1 + X2 has a normal distribution and specify its parameters.
(b) Show that X1 X2 has the same distribution as X1 + X2 .
(c) Suppose X1 and X2 are no longer independent but each still has N (0, 1) distribution. Will
X1 + X2 and X1 X2 be still independent?
c Katja Ignatieva
!
Page 41 of 43
ACTL2131 & ACTL5101
(d) Let X Gamma(, ) distributed.
1. Find the p.d.f. of an Inverse Gamma Distribution, i.e., find the p.d.f. of Y = X1 .
2. Find the c.d.f. of the inverse gamma distribution as function of the c.d.f. of the gamma
distribution.
6.
I Let Z1 and Z2 be two independent N (0, 1) random variables and let V1 2 (r1 ) and V2 2 (r2 )
be two independent chi-squared random variables. Which of the following random variables has
a t-distribution with degrees of freedom (r1 + r2 )?
Z1 + Z2
(A) 3
(V1 + V2 ) /(r1 + r2 )
Z1 + Z2
(B) 3
(V1 /r1 ) + (V2 /r2 )
Z1 + Z2
(C) 3
2 (V1 + V2 ) /(r1 + r2 )
Z1 Z2
(D) 3
(V1 + V2 ) /(r1 + r2 )
Z1
Z2
(E) 3
+3
V1 /r1
V2 /r2
II Let Z1 and Z2 be two independent standard normal random variables. Which of the following
combinations of the two has also a standard normal random variable?
(A) (Z1 + Z2 ) /2
(B)
Z1 + Z2
(C)
Z1 /Z2
(D)
Z1 Z2
(E) (Z1 Z2 ) / 2
III Let Z1 N (0, 1) and Z2 N (0, 1) be two random variables with correlation coefficient
(Z1 , Z2 ) = ,
where 1 1. Let V be a 2 (r) random variable independent of Z1 and Z2 .
Which of the following has a t-distribution with r degrees of freedom?
i.
rZ1 V 1/2
rZ V 1/2
ii.
4 2
r
iii.
(Z1 + Z2 ) V 1/2
2
4
r
iv.
(Z1 + Z2 ) V 1/2
2 ( + 1)
(A)
(B)
(C)
(D)
(E)
All
All
All
All
All
but
but
but
but
i
ii
iii
iv
IV Let X1 , X2 , . . . , Xn be i.i.d. (independent and identically distributed) Exp() random variables

.1
). Which of the following describes the distribution of the sample
(m.g.f.: MXi (t) = 1 t
mean:
n
X=
1/
Xk ?
n
k=1
(A)
(B)
(C)
(D)
(E)
X
X
X
X
X
Exp()
Exp(n)
Exp(/n)
Gamma(n, n)
.
Gamma n, n
.n
Note: m.g.f. of Gamma: MXi (t) = 1 t
).
c Katja Ignatieva
!
Page 42 of 43
ACTL2131 & ACTL5101
V Let X1 , . . . , Xn be n independent and identically distributed Poisson random variables with mean
. Describe the distribution of the sum of these random variables:
S=
n
/
Xk .
k=1
(A)
(B)
(C)
(D)
(E)
S Poisson(1)
S Poisson()
S Poisson(/n)
S Poisson(n)
Cannot be determined from the given information
VI Suppose X1 , X2 , . . . , X20 are

- twenty .independent random variables and are identically distributed
as Exp(2). Determine Pr X(20) > 1 .
.
(A) Pr X(20) > 1 = 0.94
.
(B) Pr X(20) > 1 = 0.95
.
(C) Pr X(20) > 1 = 0.96
.
(D) Pr X(20) > 1 = 0.97
.
(E) Pr X(20) > 1 = 0.98
VII Let X1 , X2 , . . . , Xn be n i.i.d. (independent and identically distributed) random variables each
with density:
fX (x) = 2x, for 0 < x < 1,
and zero otherwise.
1
2
Determine E X(n) .
1
2
(A) E X(n) = n/(n + 2)
1
2
(B) E X(n) = n/(n + 1)
1
2
(C) E X(n) = 1
1
2
(D) E X(n) = 2n/(2n + 1)
1
2
(E) E X(n) = 2n/(n + 1)
VIII In a 100-meter Olympic race, the running times are considered to be uniformly distributed between
8.5 and 10.5 seconds. Suppose there are 8 competitors in the finals. The current world record is
9.9 seconds.
Determine the probability that the loser of the race will not break the world record.
(A)
(B)
(C)
(D)
(E)
0.54
0.64
0.74
0.84
0.94
c Katja Ignatieva
!
Page 43 of 43

Week 5
1. An insurance company has a portfolio of 100 insurance contracts. The companys losses on these
contracts are independent and identically distributed. Each loss X has an exponential distribution
with mean 5 000 and each policyholder pays a premium of 5 050. Notice that each policyholder pays
an amount larger than its expected loss. Determine the probability that the aggregate loss of the
insurance company will exceed the total premiums collected. Use the normal approximation.
2. Assume that X1 , X2 , . . . , Xn is a random sample from a population with density:
5
2( x)
, for 0 < x < ;
fX (x|) =
2
0,
otherwise.
Find an estimator for using the method of moments.
3. Let X1 , X2 , . . . be a sequence of independent random variables with common mean E[Xk ] = but
different variance V ar(Xk ) = k2 . Suppose:
n
1 / 2
k 0,
n2
k=1
as n .
Prove X in probability.
4. A drunkard executes a random walk in the following manner: each minute, he takes a step north or
south, with probability 21 each, and his successive step directions are independent. Each step he takes
is of length 50 cm. Use the central limit theorem to approximate the probability distribution of his
location after one hour. Where is he most likely to be?
5. Consider N independent random variables each having a binomial distribution with parameters n = 3
and so that:
% &
3 k
nk
Pr (Xi = k) =
(1 )
,
k
for i = 1, 2, . . . , N and k = 0, 1, 2, 3, and zero otherwise. Assume that of these N random variables
n0 take the value 0, n1 take the value 1, n2 take the value 2, and n3 take the value 3 with N =
n0 + n1 + n2 + n3 .
(a) Use maximum likelihood to develop a formula to estimate .
(b) Assume that when you go to the races that you always bet on 3 races. You have taken a random
sample of your last 20 visits to the races and find that you had no winning bets on 11 visits, one
winning bet on 7 visits, and two winning bets on 2 visits. Estimate the probability of winning on
any single bet.
6. Assume that we have n independent observations y $ = [y1 , y2 , . . . , yn ], each with the Pareto p.d.f.
given by:
fYi | (yi |; A) =
A
,
yi+1
where 0 < < and 0 < A < yi < , and zero otherwise. You are now told the value of A, leaving
as the only unknown parameter.
c Katja Ignatieva
!
Page 51 of 53
ACTL2131 & ACTL5101
(a) Explain why the likelihood function L(; y, A) can be written as:
n An
,
Gn(+1)
where G = (y1 y2 . . . yn )1/n is the geometric mean of the observations.
(b) Explain why we can express the relationship between the posterior distribution, prior distribution
and likelihood function as follows:
(|y; A) fY | (y|; A)().
(c) We assume our prior pdf for is such that log() is uniformly distributed, implying:
()
1
,
0 < < .
Show that the posterior pdf for is:

(|y; A) n1 ean ,
where a = log(G/A).
(d) Explain why the posterior pdf is given by:
(|y; A) =
(an)n n1 an
e
,
(n)
0 < < .
(e) Calculate the Bayes estimator of ,

6B .
7. Using moment generating functions:
(a) show that as n , p 0 and np , the binomial distribution with parameters n and p
tends to the Poisson distribution.
(b) show that as , the gamma distribution with parameters and , properly standardised,
tends to the standard Normal distribution.
8. A random variable X with p.d.f.
fX (x) =
1
,
(1 + x2 )
for < x <
is said to have a Cauchy distribution. It is well-known that for Cauchy distribution, its mean does not
exist. Furthermore, suppose X1 , X2 , . . . , Xn are n independent Cauchy random variables, then it can
be shown that the sample mean:
n
1/
Xn =
Xk
n
k=1
also has a Cauchy distribution. Deduce then that from these results, the Cauchy violates the law of
large numbers. Explain why.
9. Given that there are n realizations of xi ,where i = 1, 2, . . . , n. We know that xi |p Ber(p) and
p U (0, 1).
(a) Find the Bayesian estimator for p.
(b) Find the Bayesian estimator for p(1 p).
(c) Why might we be interested in the Bayesian estimator for p(1 p)? Hint: consider the case when
n is large.
10. Let X1 , X2 , . . . be independent random variables with common density:
fX (x) = x(+1) ,
for x > 1,
where > 0. Define a new sequence of random variables:

Yn =
1
n1/
X(n) ,
where X(n) is the highest observation of n i.i.d. r.v. X1 , . . . , Xn .

Show that Yn converges in distribution as n and find the limiting distribution.
3 Proofs
of these results are not expected for this course.
c Katja Ignatieva
!
Page 52 of 53
ACTL2131 & ACTL5101
11. (Problem from Rice) Suppose that X1 , X2 , . . . , X20 are independent random variables with density
functions:
fX (x) = 2x, for 0 x 1,
and zero otherwise. Let S = X1 + . . . + X20 . Use the central limit theorem to approximate
Pr (S 10) .
12. Suppose that X follows a geometric distribution, with probability mass function:
Pr(X = k) = p (1 p)k1 ,
if k = 1, 2, . . ., and zero otherwise,
and assume a sample of size n.

(a) Find the method of moments estimator of p.
(b) Find the maximum likelihood estimator of p.
13. The Pareto distribution is often used in economics as a model for a density function with a slowly
decaying tail. Its density is given by:
fX (x|) = x0 x1 ,
x x0 , > 1,
and zero otherwise. Assume that x0 > 0 is given and that x1 , . . . , xn is a sample from this distribution.
(a) Find the method of moments estimate of .
(b) Find the maximum likelihood estimator of .
14. Using the p.d.f. of a chi-squared distribution with one degree of freedom:
fY (y) =
exp(y/2)
,
2y
if y > 0,
and zero otherwise, prove that the moment generating function of Y is given by:
MY (t) = (1 2t)1/2 .
15. Prove that:
d
tn1 N (0, 1) as n ,
where you might use that:
. 4
n+1
n
2.
lim
=
.
n n
2
2
16. Prove that the p.d.f. of a Snecdors F distribution, given by the transformation:
F =
U/n1
,
V /n2
where U 2 (n1 ) and V 2 (n2 ), is given by:

n /2
fF (f ) = n1 1
n /2
n2 2
((n1 + n2 )/2)
f n1 /21
.
(n1 /2) (n2 /2) (n2 + f n1 )(n1 +n2 )/2
c Katja Ignatieva
!
Page 53 of 53
Week 6

1. Let X1 , X2 , . . . , Xn be a random sample from an exponential distribution with:

fX (x|) = ex ,
x > 0,
7
and zero otherwise, where > 0. Find the value of a so that the interval from 0 to a X provides a
95% confidence interval for the parameter .
2. Consider a random sampling from a normal distribution with mean and variance 2 . Derive a
100 (1 ) % confidence interval of 2 when is known.
3. This exercise aims to show that if we sample from a continuous distribution, a pivotal quantity always
exists. Let X1 , X2 , . . . , Xn be a random sample from a continuous distribution fX (x|). Denote the
corresponding cumulative distribution function by:
* x
FX (x|) =
fX (z|) dz.
(a) Show that FX (X|) U (0, 1).

Hint: Show that Pr(FX (X|) x) = x using the quantile function (inverse of the c.d.f.),
then explain why x (representing a probability, taking values between 0 and 1) would have this
distribution.
(b) Show that W = log (1/FX (X|)) has an exponential distribution with mean 1. To do so, first
find the c.d.f c.d.f. W .
n
8
(c) From (b), deduce that
log (1/FX (Xk |)) has a Gamma distribution. Specify its parameters.
k=1
(d) Use (c) to prove that there will always be a pivotal quantity when sampling from a continuous
distribution.
4. (modified based on a past Institute of Actuaries exam.) Let X1 , X2 , . . . , Xn denote a random sample
of a Gamma(3, ) and X is the sample mean.
(a) Describe the distribution of the sample mean X.
(b) Use (a) to construct a lower 95% confidence interval for , of the form (0, U ) .
(c) Use (a) to construct an upper 95% confidence interval for , of the form (L, ).
(d) Use (a) to construct a 95% confidence interval for , of the form (L, U ) where L and U are not
necessarily equal to those found in (b) and (c).
(e) Evaluate the intervals in (b), (c) and (d) in the case for which the total of a random sample of 20
8
observations yielded a value of 20
k=1 xk = 98.2.
5. A local health club advertises that its members lose at least 10 pounds on the average during a 30-day
weight loss programme. After receiving a number of complaints from people who were enticed to join
the club, the Better Business Bureau sends out a representative to the club to check out the claim.
c Katja Ignatieva
!
Page 61 of 64
ACTL2131 & ACTL5101
The representative sampled the following nine (9) people who are enrolled in the program:
Person
1
2
3
4
5
6
7
8
9
89
xi
8i=1
9
2
i=1 xi
Before-Weight
157
174
198
205
147
165
212
169
158
1,585
283,457
After-Weight
150
167
187
198
146
153
199
171
156
1,527
262,465
Diffrence
7
7
11
7
1
12
13
-2
2
58
590
The representative of Better Business Bureau reported its findings in terms of a confidence interval.
Construct the appropriate 95% confidence interval for the average weight loss for participants in the
programme.
6. (Past Institute of Actuaries Exam
Independent
.
. random samples of size n1 and n2 are taken
- Question)
from the normal populations N 1 , 12 and N 2 , 22 . Let the sample means be X 1 and X 2 and
the sample variances be S12 and S22 . You may assume that X l and Sl2 , l = 1, 2 are independent and
distributed as follows:
%
&
2
(nk 1) Sk2
X k N k , k
and
2 (nk 1) for k = 1, 2.
nk
k2
(a) It is required to construct a confidence interval for (1 2 ), the difference between the population
means.
.
i. Suppose that 12 and 22 are known. State the distribution of X 1 X 2 and write down a
suitable pivotal quantity together with its sampling distribution. Hence, write down a 95%
confidence interval for (1 2 ).
ii. Suppose that 12 and 22 are unknown but are known to be equal. State the definition of a
tk variable in terms of independent N(0, 1) and 2k variables and use it to develop a suitable
pivotal quantity. Hence, write down a 95% confidence interval for (1 2 ).
12
, the ratio of the population variances.
22
State the definition of an Fk,l variable in terms of independent 2k and 2l variables and use it to
2
develop a suitable pivotal quantity. Hence, obtain a 90% confidence interval for 12 .
2
(b) It is required to construct a confidence interval for
(c) A regional newspaper included a consumer rights article comparing the cost of shopping in corner
shops and supermarkets. The researchers investigated the price of a standard selection of
household goods in a sample of 10 corner shops selected at random from the region, and in a
sample of 10 supermarkets selected at random from the region. The data yielded the following
values:
Sample Mean Sample S.D.
Corner Shops
22.55
1.22
Supermarkets
19.72
0.96
i. Use the result in part (a)(ii) to calculate a 95% confidence interval for (1 2 ), the difference
between the population means (1 = corner shops, 2 = supermarkets).
2
ii. Use your result in part (b) to calculate a 90% confidence interval for 12 , the ratio of the
2
population variances. Use this result to comment briefly on the assumption of equal variances
required for the confidence interval in part (c)(i).
7. (IoA, Subject CT3, April 2005, No.6) In a survey conducted by a mail order company a random sample
of 200 customers yielded 172 who indicated that they were highly satisfied with the delivery time of
their orders.
Calculate an approximate 95% confidence interval for the proportion of the companys customers who
are highly satisfied with delivery times.
c Katja Ignatieva
!
Page 62 of 64
ACTL2131 & ACTL5101
8. (IoA, Subject CT3, April 2005, No.8) The distribution of claim size under a certain class of policy is
modelled as a normal random variable, and previous years records indicate that the standard deviation
is 120.
(a) Calculate the width of a 95% confidence interval for the mean claim size if a sample of size 100 is
available.
(b) Determine the minimum sample size required to ensure that a 95% confidence interval for the
mean claim size is of width at most 10.
(c) Comment briefly on the comparison of the confidence intervals in (a) and (b) with respect to
widths and sample sizes used.
9. (IoA, Subject CT3, April 2005, No.12 (partial))

(a) A random variable Y has a Poisson distribution with parameter but there is a restriction that
zero counts cannot occur. The distribution of Y in this case is referred to as the zero-truncated
Poisson distribution.
1. Show that the probability function of Y is given by:
pY (y) =
y e
,
y!(1 e )
for y = 1, 2, 3, . . . ,
and zero otherwise.

2. Show that E[Y ] =
(b) Answer the following.
.
1 e
1. Let y1 , . . . , yn denote a random sample from the zero-truncated Poisson distribution. Show
that the maximum likelihood estimate of may be determined by the solution to the following
equation:
e
= 0,
y
1 e
and deduce that the maximum likelihood estimate is the same as the method of moments
estimate.
2. Obtain an expression for the Cramer-Rao lower bound for the variance of an unbiased estimator of .
10. (IoA, Subject 101, April 2004, No.12) For the estimation of a bernoulli probability p = Pr(success), a
series of n independent trials are performed and X represents the number of successes observed.
(a) Write down the likelihood function L(p; x) and show that the maximum likelihood estimator
(MLE) of p is p6 = X/n.
(b) Answer the following.
1. Determine the Cramer-Rao lower bound for the estimation of p.

2. Show that the variance of the MLE is equal to the Cramer-Rao lower bound.
3. Write down an approximate sampling distribution for p valid for large n.
(c) In order to develop an approximate 95% confidence interval for p for large n, the following pivotal
quantity is to be used:
p6 p
4
N (0, 1).
p(1 p)
n
Assuming that this pivotal quantity is monotonic in p, show that rearrangement of the inequality:
1.96 < 4
p6 p
p(1 p)
n
< 1.96
leads to a quadratic inequality in p, and hence determine an approximate 95% confidence interval
for p.
c Katja Ignatieva
!
Page 63 of 64
ACTL2131 & ACTL5101
(d) A simpler and more widely used approximate confidence interval is obtained by using the following
pivotal quantity
p6 p
4
N (0, 1).
p6(1 p6)
n
Determine the resulting approximate 95% confidence interval using this.
(e) In two separate applications the following data were observed:

1. 4 successes out of 10 trials
2. 80 successes out of 200 trials
In each case calculate the two approximate confidence intervals from parts (c) and (d) and comment briefly on your answers.
11. A random sample of 16 values, x1 , x2 , . . . , x16 , was drawn from a normal population and gave the
following summary statistics:
16
/
xi
51.2
x2i
243.19
i=1
16
/
i=1
Calculate a 95% confidence interval for the population mean.

12. Consider a random sample of size n from a normal distribution N(, 2 ) and let S 2 denote the sample
variance.
(a) State the sampling distribution for
for this expression when n is large.
(n 1) S 2
and specify an approximate sampling distribution
2
(b) For n = 101 calculate an approximate value for the probability that S 2 exceeds 2 by more than
a factor of 10%, i.e. Pr(S 2 > 1.1 2 ).
13. A group of 500 insurance policies gave rise to a total of 83 claims during the last year. Assuming a
Poisson model for the occurrence of claims, calculate an approximate 95% confidence interval for ,
the claim rate per policy per year.
14. Let Xi , i = 1, . . . , n denote a random sample of size n from a population with a uniform distribution
on the interval (0, ). Let X(n) = max{X1 , . . . , Xn } and define U = (1/)X(n) .
(a) Show that U has distribution function:
if u < 0;
0,
un , if 0 u 1;
FU (u) =
1,
if u > 1.
(b) Because the distribution of U does not depend on , U is a pivotal quantity. Find the 95% lower
confidence bound for .
c Katja Ignatieva
!
Page 64 of 64

Week 7
1. Explain carefully the distinction between each of the following pairs of terms:
(a) null and alternative hypotheses;
(b) one-tailed and two-tailed hypotheses;
(c) simple and composite hypotheses;
(d) Type I and Type II errors;
2. Let X1 , X2 , . . . , X10 be a random sample of size 10 from a Poisson distribution with mean . Consider
the critical region C defined by:
5
9
10
/
C = (x1 , x2 , . . . , x10 ) :
xk 3 .
k=1
(a) Show that C is a best critical region for testing H0 : = 0.1 against Ha : = 0.5.
(b) Determine the level of significance for this test.
3. Let X1 , X2 , . . . , Xn be a random sample from the density function:
%
&
1
1
2
fX (x|) = exp (x ) .
2
2
At a level of significance , find the best critical region (or most powerful test) for testing the simple
null H0 : = 0 against the simple alternative Ha : = 1.
4. Let X1 , X2 , . . . , Xn be a random sample from a Poisson() distribution. In testing the simple null
H0 : = 0 against the simple alternative Ha : = 1 , where 1 > 0 :
(a) Find the best critical region (or most powerful test).
(b) Determine the distribution of the test statistic under the null hypothesis.
5. Past Institute exam
(a) A manufacturing company produces screws of a particular size which are put into boxes of 150. On
a particular day a random sample of such boxes is taken from each of the morning and afternoon
production runs. The number of defective screws found in each sampled box are given in the
following table:
Morning
Afternoon
28
19
17
15
18
22
16
21
20
9
12
14
11
17
10
13
18
22
17
9
20
25
Table 2: Number of defectives per box

1. Test for a difference between the mean number of defectives produced in the morning and
afternoon (you may assume that the underlaying population variances are equal).
2. Plot the data in an appropriate and simple way and comment briefly on the validity of the
test of part ii).
c Katja Ignatieva
!
Page 71 of 74
ACTL2131 & ACTL5101
(b) On another day screws are put into boxes of 100. The table below gives the number of defectives
in twenty boxes sampled from this days production run.
5
6
15
18
18
14
12
9
8
18
7
12
9
11
14
5
11
18
10
12
Table 3: Number of defectives per box of 100 screws

1. Carry out a test to establish whether there is a difference between the proportions of defectives
produced on the two days.
2. Carry out a test to establish whether the proportion of defectives in boxes of 100 screws is
more than 9%.
When comparing the mean premiums for policies issued by two companies, a two-sample t test is
preformed assuming equal population variances. The sample sizes and sample variances are given by:
n1 = 25, s21 = 139.7
n2 = 30, s22 = 76.6
Preform an approximate F test at the 5% level to investigate the validity of the equal variance assumption.
The following data refers to an outbreak of botulism, a form of food poisoning that may be fatal. Each
subject is a person who contracted botulism in the outbreak. The variables recorded are the subjects
age in years, the time in hours between eating the infected food and the first signs of illness (incubation
period) and whether the subject survived (denoted by survival category Y) or died (denoted by survival
category N).
Subject
Age (x)
Incubation
period (y)
Survival
Died:
Survived:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
29 39 44 37 42 17 38 43 51 30 32 59 33 31 32 32 36 50
13 46 43 34 20 20 18 72 19 36 48 44 21 32 86 48 28 16
N Y Y N N Y N Y N N N Y N N Y N Y N
8
8 x = 405
x = 270
8
8 y = 305
y = 339
8 2
8 x2 = 15517
x = 11396
8 2
8 y 2 = 10035
y = 19665
(a) A scatterplot of incubation period against age is given below, in which different symbols are used
for subjects who died and for subjects who survived.
c Katja Ignatieva
!
Page 72 of 74
ACTL2131 & ACTL5101
A Plot of Incubation Period against age

90
Died
Survived
80
70
Incub Per
60
50
40
30
20
10
15
20
25
30
35
40
45
50
55
60
Age
Comment briefly on any relationships between age and incubation period for those subjects who
died and for those who survived.
(b) Construct suitable dotplots to investigate any relationship between:
1. age and survival, and
2. incubation period and survival
and make a brief informal comparison of the died and survived groups based on these dotplots.
(c) Construct a 95% and 99% confidence intervals for the mean difference between the incubation
period for subjects who survived and subjects who dies (i.e., take the mean incubation period for
subjects who survived minus the mean incubation period for subjects who died).
Comment briefly on these confidence intervals.
(d)
1. Construct a test to investigate whether the variances of the incubation period for subjects
who died and subjects who survived are equal.
2. Comment on the validity of the assumptions that are required for the confidence intervals
given in part c) to be approximate.

It is desired to investigate the level of premium charged by two companies for contents policies for
houses in a certain area. Random samples of 10 houses insured by company A are compared with 10
similar houses insured by company B. The premiums charged in each case are as follows:
Company A
Company B
For these data:
117
142
A = 2, 134,
154
160
8
166
166
189
188
A2 = 494, 126,
190
221
8
202
241
B = 2, 259,
233
276
8
263
279
289
284
331
302
B 2 = 541, 463.
(a) Illustrate the data given above on a suitable diagram and hence comment briefly on the validity
of the assumptions required for a two-sample t test for the premiums of these two companies.
(b) Assuming that the premiums are normally distributed, carry out a formal test to check that it is
appropriate to apply a two-sample t test to these data.
(c) Test whether the level of premiums charged by company B was higher than that charged by
company A. State your p-value and conclusions clearly.
c Katja Ignatieva
!
Page 73 of 74
ACTL2131 & ACTL5101
(d) Calculate a 95% confidence interval for the difference between the proportions of premiums of
each company that are in excess of 200. Comment briefly on your result.
(e) The average premium charged by company A in the previous year was 170. Formally test
whether company A appears to have increased its premium since the previous year.
c Katja Ignatieva
!
Page 74 of 74

Week 8
1. Explain carefully the distinction between the significance level and power.
2. Let X have a Bernoulli distribution where = Pr (X = 1). Take a random sample of size n = 10 from
this Bernoulli distribution and consider the test:
H0 : 1/2 versus Ha : > 1/2.
)
:
10
8
Using the critical region C = (X1 , X2 , . . . , X10 ) :
xk 6 :
k=1
(a) Find the power function and sketch it.
(b) Find the size of this test.

3. Recall from week 7 tutorial material question 2:
Let X1 , X2 , . . . , X10 be a random sample of size 10 from a Poisson distribution with mean . Consider
the critical region C defined by:
5
9
10
/
C = (x1 , x2 , . . . , x10 ) :
xk 3 .
k=1
Determine the power of the test under Ha .

4. Recall from week 7 tutorial material question 3:
Let X1 , X2 , . . . , Xn be a random sample from the density function:
%
&
1
1
2
fX (x|) = exp (x ) .
2
2
Determine the power of this test.
5. Prove:
nj
k /
/
j=1 i=1
(xij x) =
nj
k /
/
xj =
nj
/
xij
j=1 i=1
(xij xj ) +
k
/
j=1
nj (xj x)2
where
i=1
nj
x=
nj
k /
/
xij
j=1 i=1
k
/
nj xj
j=1
Hint: 1) Rewrite the left side by adding and subtracting within the squares xj ;
2) Rewrite is using Binomial expansion (see F&T page 2).
6. Given is that:
6=
m4 4m1 m3 + 6m2 m21 3m31

= h(m1 , m2 , m3 , m4 )
(m2 m21 )2
using the m.g.f. one can easily show that:

E[[m1 m2 m3 m4 ]$ ] = [0 1 0 3]$
Thus:
n [m1 m2 1 m3 m4 3] N (0, ).
c Katja Ignatieva
!
Page 81 of 82
ACTL2131 & ACTL5101
1
2
(a) Find = E n [m1 m2 1 m3 m4 3] [m1 m2 1 m3 m4 3]$ .
(b) Find h(m1 , m2 , m3 , m4 ).
(c) Show h(0, 1, 0, 3) = [0 6 0 1]$ .
(d) Let Xi Bin(n = 10, p = 0.2) be i.i.d. for i = 1, . . . , n. In the table below are some summary statistics of four samples. For which samples can you reject that the sample is normally
distributed? Use the chi-squared approximation.
n
x
8
xi
8i=1
n
3
i=1
8n xi4
xi
8i=1
n
2
(x
i x)
i=1
8n
(xi x)3
8i=1
n
4
i=1 (xi x)
sample 1
10
1.6
24.4
31.92
161.39
15.25
19.95
100.87
sample 2
25
2.04
34.96
14.8
140.3
17.14
7.25
68.77
sample 3
50
1.8
56
6
157.28
31.11
3.33
87.38
sample 4
100
1.92
167.36
116.22
822.76
87.17
60.53
428.52
(e) For another sample with n = 100 observations we get the value of the test statistic equals 5.4.
Find the p-value of this test.
1 2
1 2
1 2
1 2
1 2
Note
if Z1N(0,1)
then: E [Z] = 0, E Z 2 = 1, E Z 3 = 0, E Z 4 = 3, E Z 5 = 0, E Z 6 = 15,
1 7that
2
2
E Z = 0, E Z 8 = 105.
7. The following observations represent weight loss (in pounds) of men of similar physique, metabolic
activity, and so on, after a certain amount of time on three types of diet programs: A, B, and C.
Diet
A
3
7
4
5
6
-
Program
B
C
2
7
4
10
6
8
6
9
5
4
3
8
4
-
Test for the differences in the mean weight loss between the three diet programs. State any assumptions
you make. Provide the point estimates estimates of the mean losses, and the ANOVA table used to
partition the various sources of variation.
The following test concerning the mean claim amount () for a certain class of policy:
H0 : = 200 v.s. H1 : .= 200,
is to be preformed. A random sample of 50 claims is examined and yields a mean amount of 207 and
a standard deviation of 42.
Calculate the approximate p-value for the test.
c Katja Ignatieva
!
Page 82 of 82

Week 9
1. A group of 1, 725 school children were cross-classified according to their intelligence and their manner
of clothing. A result of this classification is given below:
very well clothed
well clothed
poorly clothed
dull
81
141
127
intelligent
322
457
163
very capable
233
153
48
Test for independence using a 1% level of significance.

2. (a) Past Institute exam
A 2 2 contingency table was set up to investigate whether or not two classifications criteria are
independent and resulted in the following data:
A
B
I
22
28
50
II
28
22
50
50
50
100
Calculate the observed 2 test statistic and state an approximate conclusion concerning the independence of the two criteria.
(b) (Added to the past Institute exam) Preform the Pearsons chi-square test using R.
3. Continued from previous question. Using the Fishers exact test:
(a) Write down the corresponding hypothesis and test statistic.
(b) Calculate the probability mass function of a Hypergeometric distribution with N = 100, M = 50,
n = 50 and x = 22.
(c) Use R to calculate show that the cumulative density function of a Hypergeometric distribution
with N = 100, M = 50, n = 50 and x = 22 equals 0.15867.
(d) Preform the hypothesis testing.
(e) Check your answer using R;
4. Compare the results in question 2. and 3. and explain the differences/similiraties.
A particular area in a town suffers a high burglary rate. A sample of 100 streets is taken, and in each
of the sample streets, a sample of six similar houses is taken. The table below shows the number of
sampled houses, which have had burglaries during the last six months.
No. of houses burgled
No. of streets
(a)
x
f
0
39
1
38
2
18
3
4
4
0
5
1
6
0
1. State any assumptions needed to justify the use of a binomial model for the number of sampled
houses per street which have been burgled during the last six months.
2. Derive the maximum likelihood estimator of p, the probability that a house of the type
sampled has been burgled during the last six months.
3. Fit the binomial model using your estimate of p, and, without doing a formal test, comment
on the fit.
c Katja Ignatieva
!
Page 91 of 92
ACTL2002 & ACTL5101
(b) An insurance company works on the basis that the probability of a house being burgled over a
six month period is 0.18. Carry out a test to investigate whether the binomial model with this
value of p provides a good fit for the data.
6. Check your answer of question 5b) using R.
7. Does education really make a difference in how much money you will earn?4 Researchers randomly
selected 100 people from each of three income categoriesmarginally rich, comfortably rich, and
super richand recorded their education levels. The data are summarised in the table that follows.
Highest
Education Level
No college
Some college
Undergraduate degree
Postgraduate study
Total
Marginally
Rich
32
13
43
12
100
Comfortably
Rich
20
16
51
13
100
Super Rich
23
1
60
16
100
Total
75
30
154
41
300
(a) Describe the independent multinomial populations whose proportions are compared in the 2
analysis.
(b) Provide a table with the observed proportions.
(c) Do the data indicate that the proportions in the various education levels differ for the three income
categories? Test at the = 0.01 level.
(d) Construct a 95% confidence interval for the difference in proportions with at least an undergraduate degree for individuals who are marginally and super rich. Interpret the interval.
(e) Use R to check the answer in c) and d).
4 Extended
version of [W+] 14.25
c Katja Ignatieva
!
Page 92 of 92

Week 10
1. Consider the exponential regression model with one independent variable:

x
Yi = % % i e$i
for each i = 1, 2, . . . , n,
where the +i s are independent and identically distributed normal random variables with E[i ] = 0 and
V ar(+i ) = 2 .
(a) Rewrite the exponential regression model as a linear regression model with parameters and
and describe the relationship between and % and the relationship between and % .
Derive the following from the linear regression model:
8n
8n
(xi x)2 .
(b) 6 =
ci log(yi ) where ci = (xi x)/Sxx and Sxx =
i=1
i=1
6 = x] =
(c) E[|X
6 = x] = 2 /Sxx
(d) V ar(|X
6 = x) N(1 , 2 /Sxx )
(e) (|X
(f) E[6
|X = x] =
&
x2
1
(g) V ar(6
|X = x) =
+
n Sxx
%
%
&&
x2
1
(h) (6
|X = x) N , 2
+
n Sxx
2
6 = x) = x
(i) Cov(6
, |X
Sxx
(j) What are the distributions of Y , 6% and 6% conditional on (X = x) using the LSE estimates in
the linear regression model and the relationship found in question (a)?
2. Forensic scientists use various methods for determining the likely time of death from post-mortem
examination of human bodies. A recently suggested objective method uses the concentration of a
compound (3-methoxytyramine or 3-MT) in a particular part of the brain.
In a study of the relationship between post-mortem interval and the concentration of 3-MT, samples
of the approximate part of the brain were taken from coroners cases for which the time of death had
been determined form eye-witness accounts. The intervals (x; in hours) and concentrations (y; in
parts per million) for 18 individuals who were found to have died from organic heart disease are given
in the following table. For the last two individuals (numbered 17 and 18 in the table) there was no
eye-witness testimony directly available, and the time of death was established on the basis of other
evidence including knowledge if the individuals activities.
c Katja Ignatieva
!
Page 101 of 106
ACTL2131 & ACTL5101
Observation
number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Interval
(x)
5.5
6.0
6.5
7.0
8.0
12.0
12.0
14.0
15.0
15.5
17.5
17.5
20.0
21.0
25.5
26.0
48.0
60.0
Concentration
(y)
3.26
2.67
2.82
2.80
3.29
2.28
2.34
2.18
1.97
2.56
2.09
2.69
2.56
3.17
2.18
1.94
1.57
0.61
8 2
8
8 2
8
x = 337
x = 9854.5
y = 42.98
y = 109.7936
xy = 672.8
In this investigation you are required to explore the relationship between concentration (regarded the
responds/dependent variable) and interval (regard as the explanatory/independent variable).
(a) Construct a scatterplot of the data. Comment on any interesting features of the data and discuss
briefly whether linear regression is appropriate to model the relationship between concentration
of 3-MT and the interval from death.
(b) Calculate the correlation coefficient for the data, and use it to test the null hypothesis that the
population correlation coefficient is equal to zero.
(c) Calculate the equation of the least-squares fitted regression line and use it to estimate the concentration of 3-MT:
1. after 1 day and
2. after 2 days.
Comment briefly on the reliability of these estimates.
(d) Calculate a 99% confidence interval for the slope of the regression line. Using this confidence
interval, test the hypothesis that the slope of the regression line is equal to zero. Comment on
your answer in relation to the answer given in part b) above.
3. Past Institute exam.

Consider a linear regression model in which responses Yi are uncorrelated and have expectations xi
and common variance 2 (i = 1, . . . , n), i.e. Yi is modelled as a linear regression through the origin:
E[Yi |xi ] = xi and V (Yi |xi ) = 2 (i = 1, . . . , n).
(a)
8
8
1. Show that the least squares estimator of is 61 = ni=1 xi Yi / ni=1 x2i .
2. Derive the expectation and variance of 61 under the model.
(b) An alternative to test the least squares estimator in this case is:
62 =
n
/
i=1
Yi /
n
/
xi = Y /x.
i=1
1. Derive the expectation and variance of 62 under the model.

2. Show that the variance of the estimator 62 is at least as large as that if the least squares
estimator 61 .
(c) Now consider an estimator 63 of which is a linear function of the responses, i.e. an estimator
8
which has the form 63 = ni=1 ai Yi , where a1 , . . . , an are constants.
8n
8n
1. Show that 63 is unbiased for if i=1 ai xi = 1, and that the variance of 63 is i=1 a2i 2 .
c Katja Ignatieva
!
Page 102 of 106
ACTL2131 & ACTL5101
8n
2. Show that the estimators 61 and 62 above may be expressed in the form 63 = i=1 ai Yi and
hence verify that 61 and 62 satisfy the condition for unbiasedness in c)i).
8n
3. It can be shown 8
that, subject to condition i=1 ai xi = 1, the variance of 63 is minimised by
n
setting ai = xi / i=1 x2i . Comment on this result.
4. A university wishes to analyse the performance of its students on a particular degree course. It records
the scores obtained by a sample of 12 students at the entry to the course, and the scores obtained in
their final examinations by the same students. The results are as follows:
Student
Entrance exam score x (%)
Final paper score y (%)
8
x = 836
y = 867
A
86
75
x2 = 60, 016
B
53
60
8
C
71
74
D
60
68
y 2 = 63, 603
E
62
70
F
79
75
G
66
78
H
84
90
I
90
85
J
55
60
K
58
62
L
72
70
8
(x x)(y y) = 1, 122
(a) Calculate the fitted linear regression equation of y on x.
(b) Assuming the full normal model, calculate an estimate of the error variance 2 and obtain a 90%
confidence interval for 2 .
(c) By considering the slope parameter, formally test whether the data is positively correlated.
(d) Find a 95% confidence interval for the mean finals paper score corresponding to an individual
entrance score of 53.
(e) Test whether this data come form a population with a correlation coefficient equal to 0.75.
(f) Calculate the proportion of variance explained by the model. Hence, comment on the fit of the
model.
5. Complete the following ANOVA table for a simple linear regression with 60 observations:
Source
Regression
Error
Total
D.F.
Sum of Squares
Mean Squares
F-Ratio
8.2
639.5
6. Suppose you are interested in relating the accounting variable EPS (earnings per share) to the market
variable STKPRICE (stock price). Then, a regression equation was fitted using STKPRICE as the
response variable with EPS as the regressor variable. Following is the computer output from your
fitted regression. You are also given that: x = 2.338, y = 40.21, Sx = 2.004, and Sy = 21.56.
Regression Analysis
The regression equation is
STKPRICE = 25.044 + 7.445 EPS
Predictor
Constant
EPS
Coef
25.044
7.445
SE Coef
3.326
1.144
Analysis of
SOURCE
Regression
Error
Total
Variance
DF
SS
1
10475
46
11377
47
21851
T
7.53
6.51
p
0.000
0.000
MS
10475
247
F
42.35
p
0.000
(a) Calculate the correlation coefficient of EPS and STKPRICE.

(b) Estimate the STKPRICE given an EPS of $2. Provide a 95% confidence interval of your estimate.
(c) Provide a 95% confidence interval for the slope coefficient .
(d) Compute s and R2 .
(e) Describe how you would check if the errors have constant variance.
(f) Perform a test of the significance of EPS in predicting STKPRICE at a level of significance of 5%.
c Katja Ignatieva
!
Page 103 of 106
ACTL2131 & ACTL5101
(g) Test the hypothesis H0 : = 24 against Ha : > 24 at a level of significance of 5%.

7. (Modified from an Institute of Actuaries exam problem)
An insurance company issues house buildings policies for houses of similar size in four different postcode regions A, B, C, and D. An insurance agent takes independent random samples of 10 house
buildings policies for houses of similar size in each of the four regions. The annual premiums (in
dollars) were as follows:
Region A :
247 261.
-229
8 241 270 256
8 241
x = 2, 479,
x2 = 617, 163
261
255 237.
-8 269 284 268
8 249
x = 2, 619,
x2 = 687, 467
253
229 245.
-8 247 244 245
8 221
x = 2, 441,
x2 = 597, 607
279
262 287.
-8 268 290 245
8 281
x = 2, 677,
x2 = 718, 973
Region B :
Region C :
Region D :
243 272 219

270 269 257
256 232 269
257 262 246
Perform a one-way analysis of variance at the 5% level to compare the premiums for all four regions.
State briefly the assumptions required to perform this analysis of variance.
8. You are given the following one-way ANOVA model:
Yij = + i + ij , for i = 1, . . . , I and j = 1, . . . , J
where the error terms ij are i.i.d. normal random variables with mean 0 and common variance 2 .
Using fundamental principles of maximum likelihood, derive the maximum likelihood estimates for all
parameters in the model.
9. For the one-way ANOVA model derive the following maximum likelihood estimators:
(a)
6=Y =
ni
I /
/
Yij
i=1 j=1
I
/
ni
I /
/
Yij
i=1 j=1
ni
i=1
(b)
6i = Y i. Y =
ni
/
Yij
j=1
ni
10. Suppose that Y represents a single observation from the probability density given by:
)
y 1 ,
0<y<1
fY (y|) =
0,
elsewhere.
Find the most powerful test with significance level = 0.05 to test H0 : = 2 against Ha : = 1.
11. Past Institute Exam (April 2005)
As part of an investigation into health service funding a working party was concerned with the issue of
whether mortality could be used to predict sickness rates. Data on standardised mortality rates and
standarised sickness rates collected for a sample of 10 regions and are shown in the table below:
Region
1
2
3
4
5
6
7
8
9
10
c Katja Ignatieva
!
Mortality rate m (per 100,000)

125.2
119.3
125.3
111.7
117.3
100.7
108.8
102.0
104.7
121.1
Sickness rate s (per 100,000)

206.8
213.8
197.2
200.6
189.1
183.6
181.2
168.2
165.2
228.5
Page 104 of 106
ACTL2131 & ACTL5101
Data
8 summaries:
8 2
8
8 2
8
m = 1136.1,
m = 129, 853.03,
s = 1934.2,
s = 377, 700.62, and
ms = 221, 022.58.
(a) Calculate the correlation coefficient between the mortality rates and the sickness rates and determine the probability-value for testing whether the underlaying correlation coefficient is zero
against the alternative that it is positive.
(b) Noting the issue under investigation, draw an appropriate scatterplot for these data and comment
on the relationship between the two rates.
(c) Determine the fitted linear regression of sickness rate on mortality rate and test whether the
underlaying slope coefficient can be considered to be as large as 2.0.
(d) For a region with mortality rate 115.0, estimate the expected sickness rate and calculate 95%
confidence limits for this expected rate.
12. Past Institute Exam (September 2005)
The data given in the following table are the number of deaths from AIDS in Australia for 12 consecutive
quarters starting from the second quarter of 1983.
Quarter (i)
Number of deaths (ni )
(a)
1
1
2
2
3
3
4
1
5
4
6
9
7
18
8
23
9
31
10
20
11
25
12
37
1. Draw a scatterplot of the data.

2. Comment on the nature of the relationship between the number of deaths and the quater in
this early phase of the epidemic.
(b) A statistician has suggested that a model of the form:

E[ni ] = i2
might be appropriate for these data, where is a parameter to be estimated from the data above.
She has proposed two methods for estimating , and these are given in part i. and ii. below.
812
1. Show that the least squares estimate of , obtained by minimising q = i=1 (ni i2 )2 is
given by:
812 2
i ni
.
6 = 8i=1
12
4
i=1 i
2. Show that an alternative (weighted) least squares estimate of , obtained by minimising

2 2
812
)
q = i=1 (ni i
is given by:
i2
812
; = 8i=1
12
i=1
3. Noting that
812
i=1
i4 = 60, 710 and
812
i=1
ni
i2
i2 = 650, calculate
6 and ;
for the data above.
(c) To assess whether the single parameter model which was used in part b) is appropriate for the
data, a two parameter model is considered. The model is of the form:
E[Ni ] = i
for i = 1, . . . , 12.
1. To estimate the parameters and , a simple linear regression model
E[Yi ] = + xi
is used, where xi = log(i) and Yi = log(Ni ) for i = 1, . . . , 12. Relate the parameters and
to the regression parameters and .
2. The least squares estimates of and are -0.6112 and 1.6008 with standard errors 0.4586
and 0.2525 respectively (you are not asked to verify these results).
Using the value for the estimate , conduct a formal statistical test to assess whether the
form of the model suggested in (b) is adequate.
c Katja Ignatieva
!
Page 105 of 106
ACTL2131 & ACTL5101
13. Past institute Exam

Consider the following data, which comprise of four groups sizes (y), each comprising four observations,
In scenario I, information is also given on the sum assured under the policy concerned - the sum assured
is the same for all four policies in a group. In scenario II, we regard the policies in the different groups
as having been issued by four different companies - the policies in a group are all issued the same
company.
All monetary amounts are in units of 10, 000. Summaries of the claim sizes in each group are given
in a second table.
Group
Claim sizes y
I: sum assured x
II: Company
1
0.11 0.46
0.71 1.45
1
A
2
0.52 1.43
1.84 2.47
2
B
3
1.48 2.05
2.38 3.31
3
C
4
1.52 2.36
2.95 4.08
4
D
Summaries of claim sizes:

Group
8
8 y2
y
1
2.73
2.8303
2
6.26
11.8018
3
9.22
23.0134
4
10.91
33.2289
(a) In scenario I, suppose we adopt the linear regression model

Yi = + xi + +i
where Yi is the ith claim size and xi is the corresponding sum assured, i = 1, . . . , 16.
1. Calculate the total sum of squares and its partition into the regression (model) sum of squares
and the residual (error) sum of squares.
2. Fit the model and calculate the fitted values for the first claim size of group 1 (namely 0.11)
and the last claim size of group 4 (namely 4.08).
3. Consider a test of the hypothesis H0 : = 0 against a two-sided alterative. By preforming appropriate calculations, assess the strength of the evidence against this no linear relationship
hypothesis.
(b) In scenario II, suppose we adopt the analysis of variance model
Yij = + i + eij
where Yij is the j th claim size for company i and i is the ith company effect, i = 1, 2, 3, 4 and
j = A, B, C, D.
1. Calculate the partition of the total sum of squared into the between companies (model)
sum of squares and the within companies (residual/error) sum of squares.
2. Fit the model.
3. Calculate the fitted values for the first claim size of group 1 and the last claim size of group
4.
4. Consider a test of hypothesis H0 : i = 0, i = A, B, C, D against a general alternative.
By preforming appropriate calculations, assess the strength of the evidence against this no
company effects hypothesis.
c Katja Ignatieva
!
Page 106 of 106
Week 11

1. Consider the regression model

Yk = xk + k , for each k = 1, 2, . . . , n,
that is, the regression with one regressor variable variable but without the intercept term. This model
is called regression through the origin because the true regression line passes through the point (0,0).
Derive the least squares estimate of .
Now, consider the quadratic regression model passing through the origin;
Yk = x2k + k
, for each k = 1, 2, . . . , n.
Use the previous result to determine the least squares estimate of .

2. Use the following steps to establish a relationship between the coefficient of determination and the
correlation coefficient:
(a) Show that:
y6k y = 6 (xk x) .
(b) Use this result to show that:

SSM =
n
/
k=1
where s2x is the sample variance of X.
(6
yk y) = 62 s2x (n 1) .
(c) Use the previous result to establish:

R2 = 62
s2x
= r2 .
Sy2
where s2x , s2y is the sample variance of X and Y , respectively.

3. In the regression model Yk = + xk + k , use algebra to establish the following results:
(a) R2 = 1
n 2 s2
, where s2y is the sample variance of Y .
n 1 s2y
4
n1
(b) s = sy (1 r2 )
, where sy is the sample standard deviation of Y .
n2
>
< =
6
r2
(c) t 6 = < = = n 2
1 r2
se 6
4. (a) Write down the design matrix for the simple linear regression model.
(b) Write out the matrix X $ X for the simple linear regression model.
(c) Write out the matrix X $ Y for the simple linear regression model.
(d) Write out the matrix (X $ X)1 for the simple linear regression model.
6 = (X $ X)1 X $ Y using your results above.
(e) Calculate
c Katja Ignatieva
!
Page 111 of 115
ACTL2131 & ACTL5101
5. The following model was fitted to a sample of supermarkets in order to explain their profit levels:
y = 0 + 1 x1 + 2 x2 + 3 x3 +
where
y = profits, in thousands of dollars
x1 = food sales, in tens of thousands of dollars
x2 = nonfood sales, in tens of thousands of dollars, and
x3 = store size, in thousands of square feet.
The estimated regression coefficients are given below:
61 = 0.027
and
Which of the following is TRUE?
62 = 0.097
and
63 = 0.525.
(A) A dollar increase in food sales increases profits by 2.7 cents.

(B) A 2.7 cent increase in food sales increases profits by a dollar.
(C) A 9.7 cent increase in nonfood sales decreases profits by a dollar.
(D) A dollar decrease in nonfood sales increases profits by 9.7 cents.
(E) An increase in store size by one square foot increases profits by 52.5 cents.
6. In a regression model of three explanatory variables, twenty-five observations were used to calculate
the least squares estimates. The total sum of squares and regression sum of squares were found to be
666.98 and 610.48, respectively. Calculate the adjusted coefficient of determination.
(A) 89.0%
(B) 89.4%
(C) 89.9%
(D) 90.3%
(E) 90.5%
7. In a multiple regression model given by:
y = 0 + 1 x1 + . . . + p1 xp1 + ,
which of the following gives a correct expression for the coefficient of determination?
I. SSM
SST
SST
SSE
II.
SST
III. SSM
SSE
(A) I only
(B) II only
(C) III only
(D) I and II only
(E) I and III only
8. The ANOVA table output from a multiple regression model is given below:
Source
Regression
Error
Total
D.F.
5
42
47
ANOVA Table
SS
MS
13326.1 2665.2
8525.3
203.0
21851.4
F-Ratio
13.13
Prob(> F)
0.000
Compute the adjusted coefficient of determination.

c Katja Ignatieva
!
Page 112 of 115
ACTL2131 & ACTL5101
(A) 52%
(B) 56%
(C) 61%
(D) 63%
(E) 68%
9. You have information on 62 purchases of Ford automobiles. In particular, you have the amount paid
for the car (y) in hundreds of dollars, the annual income of the individuals (x1 ) in hundreds of dollars,
the sex of the purchaser (x2 , 1 = male and 0 = female), and whether or not the purchaser graduated
from college (x3 , 1 = yes and 0 = no) . After examining the data and other information available, you
decide to use the regression model:
y = 0 + 1 x1 + 2 x2 + 3 x3 + .
You are given that:
0.109564 0.000115 0.035300 0.026804

- $ .1 0.000115
0.000001 0.000115 0.000091
X X
=
0.035300 0.000115
0.102446
0.023971
0.026804 0.000091
0.023971
0.083184
< =
and the mean square error for the model is s2 = 30106. Calculate se 62 .
(A) 0.17
(B) 17.78
(C) 50.04
(D) 55.54
(E) 57.43
10. Suppose in addition to the information in question 9., you are given:
9 558
4 880 937
.
X $Y =
7 396
6 552
Calculate the expected difference in the amount spent to purchase a car between a person who graduated from college and another one who did not.
(A) 233.5
(B) 1 604.3
(C) 2 195.3
(D) 4 920.6
(E) 6 472.1
11. A regression model of y on four independent variables x1 , x2 , x3 and x4 has been fitted to a data
consisting of 212 observations and the computer output from estimating this model is given below:
Regression Analysis
The regression equation is
y = 3894 - 50.3 x1 + 0.0826 x2 + 0.893 x3 + 0.137 x4
Predictor
Constant
x1
x2
x3
x4
Coef
3893.8
-50.32
0.08258
0.89269
0.13677
SE Coef
409.0
9.062
0.02133
0.04744
0.05303
T
9.52
-5.55
3.87
18.82
2.58
Which of the following statement is NOT true?

c Katja Ignatieva
!
Page 113 of 115
ACTL2131 & ACTL5101
(A) All the explanatory variables has a positive influence on y.

(B) The variable x1 is a significant variable.
(C) The variable x2 is a significant variable.
(D) The variable x3 is a significant variable.
(E) The variable x4 is a significant variable.
12. In a multiple regression model, which of the following gives a correct expression for the unbiased
estimate of 2 ?
=$ <
=
<
1
6
6
Y X
Y X
(A)
np+1
<
Y Y6
(B)
1
np+1
(C)
$
1
n1 Y Y
=$ <
Y Y6
=$ <
=
Y Y6
Y Y6
=$ <
=
<
1
6
6
(E) np
Y X
Y X
(D)
1
n1
<
Note: p is the rank of X.
13. The estimated regression model of fitting life expectancy from birth (LIFE EXP) on the countrys
gross national product (in thousands) per population (GNP) and the percentage of population living
in urban areas (URBAN%) is given by:
LIFE EXP = 48.24 + 0.79 GNP + 0.154 URBAN%.
For a particular country, its URBAN% is 60 and its GNP is 3.0. Calculate the estimated life expectancy
at birth for this country.
(A) 49
(B) 50
(C) 57
(D) 60
(E) 65
14. What is the use of the scatter plot of the fitted values and the residuals?
(A) to examine the normal distribution assumption of the errors
(B) to examine the goodness of fit of the regression model
(C) to examine the constant variation assumption of the errors
(D) to test whether the errors have zero mean
(E) to examine the independence of the errors
15. For the case of the multiple regression model, show:
6 = x] =
(a) E[|X
6 = x) = 2 (X $ X)1
(b) V ar(|X
16. (a) Suggest why H = X(X $ X)1 X $ is called the hat matrix.
(b) Show that HH $ = H 2 = H.
(c) Explain why:
y6i = hii yi +
hij yj ,
j'=i
where hij is the (i, j)th element of H.
c Katja Ignatieva
!
Page 114 of 115
ACTL2131 & ACTL5101
(d) Show that the (i, j)th element of H is given by:

(xi x)(xj x)
1
+
n
Sxx
for the special case of the simple linear regression model.
(e) Using H, write down an expression for the vector of residuals 6
e = Y Y6 . (Were back in the
multiple linear regression setting.)
(f) Using H, calculate E[6
e|X = x].
(g) Using H, calculate V ar(6

e|X = x).
(h) Explain why the ith standardised residual in a multiple regression model is given by e6i /(s 1 hii ),
where
E
F
n
F 1 /
e2 .
s=G
n p j=1 j
17. Use R or Excel5 to carry out the following regression calculations.
The accompanying dataset gives, for 18 countries, data on total expenditures on education as a percentage of gross national product (y), per capita income (x1 ), median educational attainment (in years)
of the population over 25 years of age (x2 ), and the ratio of the population aged 014 to the total
population (x3 ). Estimate the multiple regression model:
Y = 0 + 1 x1 + 2 x2 + 3 x3 + .
Carry out all appropriate tests to examine the significance of each predictor/regressor/independent
variable and to test for the validity of assumptions typical in a regression model.
Name of Dataset in Excel format: EDUC.xls
5 To apply linear regression in Excel you have to add a toolbox. Proceed in the following way. Go to menu in the upper left
of Excel, choose Excel options (lower right, besides Exit Excel), choose Add-Ins in the menu left, choose Go... on the
bottom line, check the box Analysis Toolpak and click on OK. Now you can go to Data, Data Analysis (just created),
Regression. Alternatively: use the functions: MINVERSE, TRANSPOSE and MMULT with Ctrl+Shift+Enter
c Katja Ignatieva
!
Page 115 of 115
Week 12

1. Consider the Excel file Exercises week 12.xlsx, tab Exercise 1.

(a) Run a linear regression and analyse the residuals?
(b) How would you model this?
(a) Use linear regression to explain the percent of successful free throws.
(b) Explain whether the variables have a direct effect or a confounding effect
(c) Check for collinearity, if collinearity is observed, how to deal with it.
(a) Explain salary using the data.
(b) Use dummy and categorical variables, which one would you prefer?
c Katja Ignatieva
!
Page 121 of 121

Sample Exam, Session 1, 2014
Sample Exam
Question 1 - [15 marks]

Suppose the joint probability density function of X1 and X2 is given by
5
x1 x2
0 x1 1, 0 x2 2,
x21 +
fX1 ,X2 (x1 , x2 ) =
3
0
otherwise.
(a)
(b)
(c)
[6 marks]
[5 marks]
[4 marks]
Find the conditional density of X2 given X1 = x1 .

Determine E[X2 |X1 = x1 ].
Verify that E [E[X2 |X1 ] = E[X2 ].

Let X be a random variable with an Erlang distribution, with parameters k = 1, 2, 3, . . . and > 0. This
distribution has the following probability density function:
fX (x) =
k xk1 ex
(k 1)!
if x > 0,
and zero otherwise.

k
(a)
[4 marks]
Show that the moment generation function is equal to MX (t) = (1 t/) .
What restrictions are there on the moment generating function in order for it to exist?
(b)
[3 marks]
Describe the relationship of the Erlang distribution with the exponential distribution. Using this relationship and the first two central moments of the exponential distribution prove
that:
k
k
E [X] =
and
V ar (X) = 2 .
Make clear what assumptions are needed to go from one step to another.
(c)
[5 marks]
We have a sample of 100 observations. From this sample we have the following
8100
8
2
information:
x
=
150
and 100
i
i=1
i=1 xi = 300. Determine the Method of Moment estimator of and
prove that the Method of Moment estimator of parameter k is exactly equal to 3.
(d)
[6 marks]
Next, we have given that the Maximum Likelihood estimator of k is 3 and
8100
x
=
150.
Determine
the
Maximum Likelihood estimator of .
i
i=1
A UNSW actuarial student will either become an fiaa or will not become an fiaa. This depends on her or
his actuarity, a peculiar mental ability which is traditionally measured by a number lying (continuously)
between zero and one. Without testing, we do not know the actuarity of any particular person; we do know
that students with higher actuarity coefficients are more likely to become actuaries. As a result of painstaking
research by UNSW staff, it has been established that a reasonable prior distribution for actuarity is Beta
with parameters (, ).
Researchers have also established that, given a students actuarity , her or his ability to become an
actuary is Bernoulli distributed,
(I|) Bernoulli().
(a)
(b)
(c)
[5 marks] For this model, what is the marginal distribution of I?

[10 marks] What is the posterior density, (|i)?
[5 marks] Calculate the Bayesian estimator of .
Question 4 - [18 marks] The table below gives 105 observations of the times (in minutes) between calls
to the emergency services telephone number. These times have been grouped into nine minute intervals (the
left-hand column, Class).
c Katja Ignatieva
!
Page 131 of 132
ACTL2131 & ACTL5101
Class
A1 = [0, 9]
A2 = (9, 18]
A3 = (18, 27]
A4 = (27, 36]
A5 = (36, 45]
A6 = (45, 54]
A7 = (54, 63]
A8 = (63, 72]
A9 = (72, )
Frequency
41
22
11
10
9
5
2
3
2
Probability
0.362
0.231
0.147
0.094
0.060
0.038
0.024
0.016
0.027
Sample Exam, Session 1, 2014
Expected
38.052 0
24.265 5
15.466 5
9.870 0
6.289 5
4.011 0
2.562 0
1.627 5
2.866 5
(a)
[10 marks] A colleague suggests to you that the distribution of times is exponential with a mean
of 20. Test whether this is the case.
(b)
[5 marks] Above, you have assumed a mean of 20. If, instead, you had to estimate the mean, in
what way would your testing procedure differ?
(c)
[3 marks] Suggest some graphical ways of investigating whether these data are exponential with
mean 20.
Aggregate national data for a particular industry are provided below. You have been asked to fit a CobbDouglas production function to these data.
Country
Output, Q
Labour, L
Capital, K
B
80
60
50
C
150
100
100
D
135
100
80
E
165
120
100
F
95
70
60
G
130
90
80
H
110
80
70
(a)
[1 mark] A Cobb-Douglas production function takes the form Q = AL K . How will you fit
this model to the data?
(b)
[5 marks] For the model you have chosen to fit, you are given (in standard notation) that
62.780 946 07 55.746 515 58 24.179 864 75

(X$ X)1 = 55.746 515 58 204.981 847 2 181.993 220 7
24.179 864 75 181.993 220 7 175.080 437 9
7
13.559 763 03 13.128 399 27
X$ X = 13.559 763 03 26.330 018 58 25.496 918 45
13.128 399 27 25.496 918 45 24.696 210 72
14.5600586
X$ Y = 28.27258881
27.38094584
Determine 6 .
(c)
[5 marks] Given
/
yi2 = 30.360 112 93
y6i2 = 30.359 956 6
y = 2.080 008 372
calculate SST and SSM.

(d)
[3 marks] If the SSE were 0.000 156 327, give estimates of the variance of
6 and 6 and their
covariance.
(e)
[5 marks] Carefully detailing your steps, test for the significance of and at the 5% level.
(f)
[2 marks] Calculate the adjusted R2 of your model.
(g)
[4 marks] Test whether your model is homogeneous of degree one, that is, whether + = 1,
at a 5% level of significance.
-End of Sample Exam-
c Katja Ignatieva
!
Page 132 of 132

Additional material

Additional material
Includes:
- UNSW TV: Back to Basics: Basic Mathematical Tools for Actuarial Students
- Greek alphabet
- List of some Excel functions
- Mathematics (Integrals, differentiation & mathematic rules)
- Linear Algebra
- Formulae and Tables
Session 1, 2014
ACTL2131 & ACTL5101
UNSW TV&Greek alphabet
UNSW TV: Back to Basics: Basic Mathematical Tools for Actuarial Students
Back to Basics: List of all videos
http://tv.unsw.edu.au/collection/BBA28340-2B1D-11DF-AA9D123139020041/mediaType/unswVideo
Back to Basics: Introduction
http://tv.unsw.edu.au/video/back-to-basics-introduction
Back to Basics: Topic 1: Sum of geometric series

http://tv.unsw.edu.au/video/back-to-basics-topic-1
Back to Basics: Topic 2: Exponential and logarithm function
Back to Basics: Topic 3: Quadratic formula
Back to Basics: Topic 4: Taylor series
Back to Basics: Topic 5: Derivative
Back to Basics: Topic 6: The product rule, the quotient rule and the chain rule (derivatives)
Back to Basics: Topic 7: Integration by parts
Back to Basics: Topic 8: Differential of an integral
Back to Basics: Topic 9: Newton-Ralphson approximation
Greek alphabet
Name
Lower
case
alpha
beta
gamma
delta
epsilon +,
zeta
eta
theta
,
iota
kappa
lambda
mu
c Katja Ignatieva
!
Upper
case
A
B
E
Z
H
I
K
Name
nu
xi
omikron
pi
rho
sigma
tau
upsilon
phi
chi
psi
omega
Lower
case
o
, 1
, 2
,
Upper
case
N
Page 1002 of 1113
ACTL2131 & ACTL5101
List of general Excel functions:

Function
Explanation
exp(1)
returns e
List of some Excel functions
Function
correl(a1:a4,b1:b4)
ln
natural log
covar(a1:a4,b1:b4)
pi()
returns pi
var(a1:a4)
exp(gammaln(alpha))
gamma function evaluated in alpha
stdev(a1:a4)
fact(a1)
returns the factorial of

cell a1
returns n choose r
skew(a1:a4)
combin(n,r)
sqrt(a1)
returns the square root

of cell a1
sum(a1:b4)
returns the sum of the

cells A1-B4
returns the sum of the
squared values of cells
A1-B4
returns the sum of
product a1b1-a4b4
returns the average of
array A1-A4
returns the alpha-th
trimmed mean of the
array A1-A4
returns the median of
the array A1-A4
counts the number of
cells with a number in
the array A1-A4
counts the number of
cells equal to alpha in
the array A1-A4
returns the alpha(
1)-th percentile of the
array A1-A4
returns the alpha(=
0, 1, 2, 3, 4)-th quartile
of the array A1-A4
sumsq(a1:b4)
sumproduct(a1:a4,b1:b4)
average(a1:a4)
trimmean(a1:a4,alpha)
median(a1:a4)
count(a1:a4)
countif(a1:a4,alpha)
percentile(a1:a4,alpha)
quartile(a1:a4,alpha)
Function
if(test,value true, value false)
and(test1, test2, test3)
rand()
randbetween(a,b)
vlookup(a1,a1:c4,r)
hlookup(a1,a1:c4,r)
c Katja Ignatieva
!
kurt(a1:a4)
abs(a1)
round(a1,r)
floor(a1,r)
ceiling(a1,r)
min(a1:a4)
max(a1:a4)
Explanation
returns the correlation
coefficient between array A1-A4 and B1-B4
returns the covariance
between array A1-A4
and B1-B4
returns the variance of
array A1-A4
returns the standard
deviation of array A1A4
returns the skewness of
array A1-A4
returns the kurtosis of
array A1-A4
returns the absolute

value of cell A1
rounds using r digits
rounds down using r

digits
rounds up using r digits
returns the minimum
of array A1-A4
returns the maximum
of array A1-A4
Explanation
produces value true if condition is met and value false otherwise
produces only TRUE if all conditions are met, otherwise FALSE
returns a random number between 0 and 1
returns a random number between a and b
looks for value of cell a1 in the array A1-A4 and returns the r-th
column
looks for value of cell a1 in the array A1-C1 and returns the r-th
row
Page 1003 of 1113
ACTL2131 & ACTL5101
List of some Excel functions
List of statistical Excel functions:

Be aware of the differences in the definition of distributions and check which definition it is, and adjust your
parameters to it.
Continuous distributions
Discrete distributions
Function
Explanation
Function
Explanation
normdist(x,mu,sd,1)
returns normal cdf
poisson(x,lambda,1)
returns poisson cdf
normdist(x,mu,sd,0)
returns normal pdf
poisson(x,lambda,0)
returns poisson pmf
norminv(p,mu,sd)
returns the p-th quanPoisson inverse
make a table with the
tile of a normal r.v.
poisson c.d.f. and you
if function to find the
inverse?
lognormdist(x,mu,sd)
loginv(p,mu,sd)
returns Lognormal cdf

returns the p-th quantile of a Lognormal r.v.
binomdist(x,n,p,1)
binomdist(x,n,p,0)
returns Binomial cdf

returns Binomial pmf
critbinom(alpha,n,p)
returns the alpha-th

quantile of a Binomial
r.v.
returns Negative Binomial pmf
use relationship between NB and GEO
NBin c.d.f. and you if
function to find the inverse?
expondist(x,lambda,1)
returns exponential cdf
negbinomdist(x,n,p)
expondist(x,lambda,0)
returns
pdf
Geometric distribution
exponential
use relation exponential

and gamma distribution
to find the p-th quantile of
an exponential r.v
NBin inverse
gammadist(x,alpha,beta,1)
returns gamma cdf
hypgeodist(x,n,M,N)
gammadist(x,alpha,beta,0)
returns gamma pdf
Hypgeometric inverse
gammainv(p,alpha,beta)
returns the p-th quantile of a gamma r.v.
weibull(x,alpha,beta,1)
weibull(x,alpha,beta,0)
returns weibulll cdf

returns weibull pdf
betadist(x,alpha,beta)
betainv(p,alpha,beta)
returns beta cdf

returns the p-th quantile of a beta r.v.
chidist(x,df)
chiinv(p,df)
returns chi-squared cdf

returns the p-th quantile of a chi-squared
r.v.
tdist(x,df)
tinv(1-p/2,df)
returns student-t cdf

returns the 1-p-th
quantile of a student-t
r.v.
fdist(x, df1,df2)
finv(p, df1,df2)
returns F cdf
returns the p-th quantile of a F r.v.
c Katja Ignatieva
!
returns Hypergeometric cdf

Hypgeometric
c.d.f.
and you if function to
find the inverse?
Page 1004 of 1113
ACTL2131 & ACTL5101
Required Pre-Knowledge Mathematics

Required Pre-knowledge Mathematics
List of integrals
Hb
a
Hb
a
Hb
a
Hb
Hb
c xd dx
a
Hb c
a dx+e dx
Jb
c
d+1
,
d+1 x
a
1c
d log (d x +
I
2b
e) a ,
b
[x log (c x) x]a ,
I
Jb
exp(cx)
,
c
a
Jb
I
exp(cx)
(c
1)
,
2
c
a
I
< 2
exp (c x) xc 2x
c2 +
=
log (c x) dx =
exp (c x) dx =
x exp (c x) dx =
x2 exp (c x) dx =
Hb
sin (x) dx =
H ab
cos
(x) dx =
H 1 a
x
exp
(x)
dx =
H 10 1
1
x
(1
x)
dx =
0 H
- 2 .
exp x /2 dx =
0
a
b
[ cos (x)]a
[sin (x)]ba
f (x)dx
a
! b b
a f (x)dx
a
for a, b, c (, c .= 0;
2
c3
=Jb
()()
(+)
3
/2
Hb
H b dv
b
u dx = [u v]a a v du
dx dx,
H ba dx
Ha
f
(x)dx
=
f
(x)dx,
a
b
Hb
H g(b)
%
f
(g
(x))
g
(x)
dx
=
f (x) dx,
a
g(a)
!b
for a, b, c, d, e (, d .= 0;
for a, b, c (, c .= 0;
()
Integral rules
for a, b, c, d (, d .= 1;
for a, b, c (, c .= 0;
, for a, b, c (, c .= 0;
for a, b (;
for a, b (;
for > 0, week 2 Prob & Stat;
for , > 0, week 2 of Prob & Stat;
week 2 of Prob & Stat (Gaussian).
Integration by parts (see page 3 F&T);

Reversing limits;
Integration by substitution;
= f (b);
= f (a),
Differentiation of integral.
Differentiation rules
f (x)g(x)
<x =
f (x)
x g(x)
f (g(x))
x
c Katja Ignatieva
!
=
=
=
f (x)
g(x)
x g (x) + f (x) x ,
g(x)
f (x)
g(x)f
(x)
x
x
,
(g(x))2
%
%
f (g (x)) g (x) ,
Product rule;
Quotient rule, g (x) .= 0;
Chain rule.
Page 1005 of 1113
ACTL2131 & ACTL5101
Required Pre-Knowledge Mathematics
Mathematical rules
-nx!. =
k =
exp (x) =
log (1 + x) =
n
(a + b) =
f8
(x + h) =
n
8nk=0 x2 =
k=0 x =
8
k
k=0 x =
8n1
k
k=0 x =
8
k
x =
8k=n 1
k=1 k2k =
exp (a + b) =
exp (a b) =
n
exp (a) =
log (a b) =
log (a/b) =
log (an ) =
2
ax + bx + c = 0
c Katja Ignatieva
!
x (x 1)!,
0! = 1,
n!
,
(nk)!k!
8
xi
i=0 i! ,
8
i+1 xi
i ,
-n. i ni
8ni=1 (1)
a b ,
i=0
i
8 hi (i)
(x),
i=0 i! f
n(n + 1)/2
n(n + 1)(2n + 1)/6
1
1xn
1x
1x
xn
1x
log(2)
exp (a) exp (b)
exp (a) / exp (b)
exp (n a)
log (a) + log (b)
log (a) log (b)
n log (a).
b2 4ac
x = b 2a
Factorial function, x N+ = {0, 1, 2, . . .};

Binomial coefficient, n, k N+ ;
Exponential function (see page 2 F&T);

Natural log function, 1 < x 1 (see page 2 F&T);
Binomial expansion, n N+ (see page 2 F&T);
Taylor series expansion (see page 3 F&T);
for n N ;
for n N ;
Geometric series, for 1 < x < 1;
abc formula.
Page 1006 of 1113

Required Pre-knowledge Linear Algebra
Terminology
Matrix
We denote a matrix:
A=
A11
A21
..
.
A12
A22
..
.
...
...
..
.
A1m
A2m
..
.
An1
An2
. . . Anm
where Ai,j is the (i, j)-element of the matrix A, i.e., the ith row, j th column element of the matrix A. The
matrix is said to have size (n m), i.e., n rows and m columns.
Vector
We denote a vector:
x=
x1
x2
..
.
xn
where xi is the ith element of the vector x. The length of the vector is n, i.e., the number of elements of the
vector. The size is n 1, i.e., a vector is a matrix with only one column
Transpose
The transpose of a matrix A with size (n m) is given by:
A11
A11 A21 . . . An1
A21
A12 A22 . . . An2
where A = .
A$ = .
.. ,
..
..
..
..
.
.
.
An1
A1m A2m . . . Anm
A12
A22
..
.
...
...
..
.
A1m
A2m
..
.
An2
. . . Anm
Thus the A$ (which is also denoted by A1) is a matrix of size (m n) where the (i,j) element of the matrix
A$ is the (j,i) element of the matrix A.
Matrix multiplication
Vector and vector
Multiplication of two vectors x and y with same length n:
x$
[1n][n1]
n
/
xi yi ,
i=1
[11]
where [n m] denotes n row and m columns.

Note: this vector multiplication is only possible if vectors x and y have the same length (number of
elements). The result of the product of the row vector x$ and vector y is a constant/number R or a
1 1 matrix.
ACTL2131 & ACTL5101
Required Pre-Knowledge Linear Algebra
Matrix and matrix

Multiplication of two matrices A with size (n k) and B with size (k m):
A
k
/
Ail Blj ,
[nk][km]
C ,
[nm]
where
Cij =
l=1
hence, the (i, j) element of the matrix C is given by multiplying the ith row of A with the j th column of B.
Note: number of columns of matrix A should be equal to the number of rows of matrix B. The matrix
C has the number of rows equal to the number of rows of matrix A and the number of columns of matrix C
is equal to the number of columns of matrix B.
Matrix and vector
Multiplication of a matrix A with size (n m) with a vector x with length m (i.e., size (m 1):
A
[nm][m1]
y
[n1]
where
yi =
m
/
Ail xl ,
l=1
hence, the ith element of the vector y is obtained by multiplying the ith row of the matrix A with the vector
x.
Note that the number of columns of the matrix A should be equal to the length of the vector x. The
length of vector y will be equal to the number of rows of vector A.
Vector and vector II
In Section 3 we multiplied two vectors (a row vector with a column vector) which resulted in a constant.
Alternatively, one can multiply a column vector with a row vector to obtain a matrix. Note that this
does not occur often.
Alternative multiplication of two vectors x with length n and y with length m:
x
y$ =
[n1][1m]
A ,
[nm]
where
Aij = xi yi ,
hence, the (i, j) element of A is the product of the ith element of vector x and the j th element of vector y.
Matrix and solving system of equations

For solving the following system of equations of n equations with m parameters:
a11 x1 + a12 x2 + . . . + a1m xm = y1
a21 x1 + a22 x2 + . . . + a2m xm = y2

..
..
..
..
..
.
.
.
.
.
an1 x1 + an2 x2 + . . . + anm xm = yn
we can use linear algebra. Denoting in matrices we have:

A
[nm][m1]
c Katja Ignatieva
!
y ,
(1)
[n1]
Page 1008 of 1113
ACTL2131 & ACTL5101
where,
A=
A11
A21
..
.
A12
A22
..
.
...
...
..
.
A1m
A2m
..
.
An1
An2
. . . Anm
,x =
x1
x2
..
.
xm
, and y =
y1
y2
..
.
yn
To solve equation (1), using n = m (i.e., number of equations equals the number of parameters) we proceed
as follows:
1. Create the matrix:
[A | y] =
A11
A21
..
.
A12
A22
..
.
...
...
..
.
A1n
A2n
..
.
y1
y2
..
.
An1
An2
. . . Ann
yn
2. Using elementary row operations (see below) create
1 0
0 1
[In | z] = . .
.. ..
0
a matrix:
... 0
... 0
.
..
. ..
0 ... 1
where In is an (n n) identity matrix.
z1
z2
..
.
zn
3. Then, we have that x = z.

Elementary row operations
Let A be a (n m) matrix. Let Ri be the ith row of the matrix A. Then, the elementary row operations
are:
1. Row switching:
A row within the matrix A

R1
R1
.. ..
. .

Ri Rj

A = ... ...

Rj Ri

. .
.. ..
Rn
Rn
can be switched with another
A11 A12 . . .
..
..
..
.
.
.
Ai1 Ai2 . . .
..
..
A = ...
.
.
Aj1 Aj2 . . .
.
..
..
..
.
.
An1
2. Row multiplication (c .= 0):

Each element of a row can be multiplied with
R1
R1
A11
.. ..
..
. .
.
A = Ri c Ri
A=
Ai1
. .
.
.
.
. .
..
Rn
Rn
An1
An2
row in the matrix:

A11 A12
A1m
..
.. ..
.
.
.
Aj1 Aj2
Aim

..
.. ..
.
.
.
Ajm
Ai1 Ai2
.
..
..
.
. ..
An1 An2
. . . Anm
a non-zero constant c R\{0}:

A12 . . . A1m
A11 A12
..
.. ..
..
..
.
.
.
.
.
cAi1 cAi2
Ai2 . . . Aim

..
.. ..
..
..
.
.
. .
.
An2 . . . Anm
An1 An2
...
..
.
...
..
.
A1m
..
.
Ajm
.. .
.
Aim
..
.
...
..
.
. . . Anm
. . . A1m
..
..
.
.
. . . cAim
..
..
.
.
. . . Anm
3. Row addition:
c Katja Ignatieva
!
Page 1009 of 1113
ACTL2131 & ACTL5101
A row can be replaced by the row itself plus another row:
R1
R1
..
..
.
Ri
Ri
..
.
..
A = .
Rj Rj + Ri
..
..
Rn
A11
..
.
Ai1
= ...
Aj1
.
..
An1
Rn
A12
..
.
Ai2
..
.
Aj2
..
.
An2

A1m
A11
..
..
.
.

Aim
A
i1

..
..
.
.

Ai1 + Aj1
. . . Ajm

..
..
..
.
.
.
. . . Anm
An1
...
..
.
...
..
.
Note, combining all three we have:
R1
Ri + cRj
..
..
.
Ri
R
1
..
.
..
A = .
Rj
Rj
.
..
..
Rn
Rn

A11 A12 . . . A1m
Ai1 + Aj1
..
..
.
..
.
..
..
.
.
.

Ai1 Ai2 . . . Aim
A
11

..
.
.
.
.
..
..
..
..
= .

Aj1 Aj2 . . . Ajm
Aj1

.
..
..
..
.
.
.
.
.
.
.
.
An1 An2 . . . Anm
An1
A12
..
.
...
..
.
...
..
.
Ai2
..
.
Ai2 + Aj2
..
.
An2
Ai2 + Aj2
..
.
Aim
..
.
. . . Aim + Ajm
..
..
.
.
...
Anm
Aj2
..
.
. . . Aim + Ajm
..
..
.
.
...
A1m
..
..
.
.
...
Ajm
..
..
.
.
An2
...
A12
..
.
A1m
..
.
Anm
Example Gaussian Elimination

In a matrix:
- a leading row is one which is not all zeros;
- in a leading row, the leading entry is first (leftmost) non-zero entry;
- a leading column contains the leading entry for some row.
Example:
1 2
0 2
0 0
3
4
1
In this case, elements a11 , a22 , and a33 are leading entries, thus all rows are leading rows and all columns
are leading columns.
Gaussian elimination: Given an augmented matrix C = [A|y] with n rows/equations. Solve:
3x1 + 2x2 =1
2x1 1x2 =3
c Katja Ignatieva
!
Page 1010 of 1113
ACTL2131 & ACTL5101
The augmented matrix is:

+
3
2
2 1
1 3
Then solve it by:

1. Multiply first row by 1/3 R1 R1 /3;
2. Subtract twice the (new) first row from the second row R2 R2 2R1
Hence we now have:
+
3
2
2 1
1 3
3/3
2/3
1/3
2 2 1 4/3 3 2/3
1
0
2/3 1/3
7/3 7/3
The next steps are:

1. Multiply second row by -3/7 R2 R2 / 7/3;
2. Subtract 2/3 the (new) second row from the first row R1 R1 2/3R2
Hence we now have:
+
, +
1 2/3 1/3
1
0 7/3 7/3
0
2/3 1/3
1
1
1
0
2/3 2/3 1/3 (2/3)

1
1
1
0
0 1
1 1
Thus we have that x1 = 1 and x2 = 1.
Inverse Matrices
An inverse of a matrix A only exists for square matrices (i.e., the size of matrix A is n n). For non-singular
square matrices (i.e., matrices with an inverse) we have the following:
A1 A = In
and we have: AA1 = In ,
where In is an identity matrix of size n n.

For 2 2 matrices
Denote the matrix 2 2 matrix A:
A=
a
c
b
d
Then, the inverse of the matrix A is obtained by:

1
1
=
det(A)
d b
c a
where det(A) = a d b c. Note that the matrix A is non-singular if and only if det(A) .= 0. Hence, if
det(A) = 0 then the matrix A is singular and its inverse does not exist, which can be observed from the fact
that you cannot divide by zero.
For 3 3 matrices
Denote the matrix 3 3 matrix A:
a b
A= d e
g h
c
f .
i
Then, the inverse of the matrix A is obtained by:
(ei f h) (ch bi) (bf ce)

1
(f g di) (ai cg) (cd af ) ,
A1 =
det(A)
(dh eg) (gb ah) (ae eg)
where det(A) = a (ei f h) b (id f g) + c (dh eg). Note that the matrix A is non-singular if and only
if det(A) .= 0. Hence, if det(A) = 0 then the matrix A is singular and its inverse does not exist, which can
be observed from the fact that you cannot divide by zero.
c Katja Ignatieva
!
Page 1011 of 1113
ACTL2131 & ACTL5101
For n n matrices
To find the inverse of an n n matrix A consists of the following steps:
1. Create the following n 2n matrix:
[A In ] =
A11
A21
..
.
A12
A22
..
.
...
...
..
.
A1n
A2n
..
.
1
0
..
.
An1
An2
. . . Ann
0 ... 0
1 ... 0
.. . .
.
. ..
.
0 ... 1
2. Using Gaussian elimination (also referred to as Gauss-Jordan) (i.e., elementary row operations, see
section 3) create the matrix [In B].
3. Now, the matrix B is the inverse of the matrix A, i.e., B = A1 .
Blockwise inversion
Let X be a symmetric (n + m) (n + m) matrix, A be a symmetric n n, D be a symmetric m m, B be
a n m matrix, and C be a m n matrix, then we have:
X 1 =
A
C
B
D
,1
A1 + A1 B(D CA1 B)1 CA1

(D CA1 B)1 CA1
A1 B(D CA1 B)1

(D CA1 B)1
Eigenvalues and Eigenvectors

Let A be a symmetric n n matrix, -a scaler- the eigenvalue, c -a vector- the eigenvector if:
Ac = c.
In general, there are n solutions 1 , . . . , n called the eigenvalues (characteristic roots) of A, corresponding
with n eigenvectors c1 , . . . , cn called the eigenvectors (characteristic vectors)
A matrix is positive definite if all its eigenvalues are positive. A matrix is positive semi-definite if all its
eigenvalues are non-negative. The determinant of a symmetric matrix equals the product of its n eigenvalues.
The rank of a symmetric matrix corresponds to the number of non-zero eigenvalues. If an eigenvalue is zero,
the matrix A is not of full rank and thus singular.
Differentiation
Let x be an n-dimensional column vector, c be an n-dimensional column vector, then c$ x is a scalar.
Consider:
c$ x
= c.
x
This is a column vector of n derivatives the typical element being ci . More generally, we have for a vectorial
function Ax (where A is a matrix) that:
Ax
= A$ .
x
The element in column i, row j in this matrix is the derivative of the j th element in the function Ax with
respect to xi . Further:
x$ Ax
= 2Ax
x
for a symmetric matrix A. If A is not symmetric, we have
x$ Ax
= (A + A$ )x.
x
All these results follow from collecting the results from an element-by-element differentiation.
c Katja Ignatieva
!
Page 1012 of 1113
ACTL2131 & ACTL5101
Least Squares Manipulations

Let xi = [xi1 , xi2 , . . . , xiK ]
with xi1 1 and = [1 , 2 , . . . , K ] , then:

x$
i = 1 + 2 xi2 + . . . + K xiK .
The matrix:
N
/
i=1
xi x$
i =
N
/
i=1
xi1
xi2
..
.
xiK
8N
8N
i=1
..
.
..
.
[xi1 , xi2 , . . . , xiK ]
8N
x2i1
i=1
xi2 xi1
8N
i=1 xi1 xiK
i=1
...
x2i2
..
.
...
8N
i=1 xi2 xiK
8N
xiK xi1
..
.
..
.
8N 2
i=1 xiK
i=1
is a K K symmetric matrix containing sums of squares and cross-products. The vector:
has length K, so that the system:
8N
i=1 xi1 yi
N
8N xi2 yi
/
i=1
xi yi =
..
.
i=1
8N
i=1 xiK yi
KN
/
xi x$
i
i=1
N
/
xi yi
i=1
is a system of K equations with K unknowns. With matrix notation, the N K matrix X is defined as:
x11 x12 . . . x1K
..
..
..
X = ...
.
.
.
xN 1
xN 1
...
xN K
and y = [y1 , y2 , . . . , yN ] . From this it can easily be verified that:

X $X =
N
/
xi x$
i
i=1
X $y =
N
/
xi yi
i=1
c Katja Ignatieva
!
Page 1013 of 1113
1.2
1.3
Solving Equations
Newton-Ralphson method
If x is a sufficient good approximation to a root of the equation
f (x) = 0 then (provided convergence occurs) a better approximation
is
x) = x ff!(x)
(x)
Integrating factors
The integrating factor for solving the differential equation
dy
(x)y = Q(x)
dx + P-H
. is:
exp P (x)dx
Second order difference equations
The general solution of the difference equation
axn+2 + bxn+1 + cxn = 0 is:
if b2 4ac > 0: xn = An1 + Bn2
(distinct real roots, 1 .= 2 )
if b2 4ac = 0: xn = (A + Bn)n
(equal real roots, 1 = 2 = )
if b2 4ac < 0: xn = rn (A cos n + B sin n)
(complex roots, 1 = 2 = rei )
where 1 and 2 are the roots of the quadratic equation
a2 + b + c = 0
1.4
1.5
Calculus
Taylor series (one variable)
2
f (x + h) = f (x) + hf % (x) + h2! f %% (x) + . . .
Taylor series (two variables)
f (x +
y + k) = f (x, y) + hfx% (x, y) + kfy% (x,.y)
- h,
1
%%
%%
%%
+ 2! h2 fxx
(x, y) + 2hkfxy
(x, y) + k 2 fyy
(x, y) + . . .
Integration by parts
H b dv
Hb
b
u dx = [uv]a a v du
dx dx
a dx
Double integrals (changing
the order= of integration)
H b -H x
.
H b <H b
f (x, y)dy dx = a y f (x, y)dx dy or
a
a
Hb Hx
Hb Hb
a dx a dyf (x, y) = a dy y dxf (x, y)
The domain of integration here is the set of values (x, y) for which
a y x b.
Differentiating an integral
H b(y)
d
%
%
dy a(y) f (x, y)dx = b (y)f [b(y), y] a (y)f [a(y), y]
H b(y)
f (x, y)dx
+ a(y) y
Gamma function
DefinitionH
(x) = 0 tx1 et dt, x > 0

Properties
(x) = (x 1)(x 1)
(n) = (n 1)!, n = 1, 2, 3, . . .
(1 /2 ) =
Bayes formula
Let A1 , A2 , . . . , An be a collection of mutually exclusive and exhaustive
events withP (Ai ) .= 0, i = 1, 2, . . . , n.
For any event B such that P (X) .= 0:
P (B|Ai )P (Ai )
P (Ai |B) = "
, i = 1, 2, . . . , n.
n
P (B|Aj )P (Aj )
j=1
Formulae and Tables
Page 1100 of 1113
Mathematical Methods
Series
Exponential function
2
3
exp(x) = ex = 1 + x + x2! + x3! + . . .
Natural log function
3
2
(1 < x 1)
log(1 + x) = ln(1 + x) = x x2 + x3 . . .
Binomial expansion
- .
- .
(a + b)n = an + n1 an1 b + n2 an2 b2 + . . . + bn
where n is a positive integer
x2 + p(p1)(p2)
x3 + . . .
(1 + x)p = 1 + px + p(p1)
2!
3!
(1 < x 1)
ACTL2131 & ACTL5101
c Katja Ignatieva
!
1
1.1
2.1
Statistical distributions
Notation
PF
=Probability function, p(x)
PDF
=Probability density function, f (x)
DF
=Distribution function, F (x)
PGF
=Probability generating function, G(s)
MGF
=Moment generating function, M (t)
Note. Where formulae have been omitted below, this indicates that
(a) there is no simple formula or (b) the function does not have a
finite value or (c) the function equals zero.
Discrete distributions
Binomial distribution
Parameters: n, p (n=positive
integer, 0 < p < 1 with q = 1 p)
- .
PF:
p(x) = nx px q nx , x = 0, 1, 2, . . . , n
DF:
The distribution function is tabulate in the statistical
tables section.
PGF:
G(s) = (q + ps)n
MGF:
M (t) = (q + pet )n
Moments:
E[X] = np, var(X) = npq
Coefficient
of skewness: qp
npq
Moments:
Coefficient
of skewness:
E[X] = kp , V ar(X) =
2p
kq
kq
p2
Negative binomial distribution - Type 2

Parameters: k, p (k > 0, 0 < p < 1 with q = 1 p)
(k+x)
PF:
p(x) = (x+1)(k)
pk q xk , x = 0, 1, 2, . . .
=k
<
p
PGF:
G(s) = 1qs
=k
<
p
MGF:
M (t) = 1qe
t
kq
Moments:
E[X] = kq
p , V ar(X) = p2
Coefficient
2p
of skewness:
kq
Geometric distribution
The geometric distribution is the same as the negative binomial
distribution with parameter k = 1.
Formulae and Tables
Page 1101 of 1113
Negative binomial distribution - Type 1

Parameters: k, p (k=positive
integer, 0 < p < 1 with q = 1 p)
- . k xk
PF:
p(x) = x1
p
q
, x = k, k + 1, k + 2, . . .
<k1 =k
ps
PGF:
G(s) = 1qs
< t =k
pe
MGF:
M (t) = 1qe
t
Bernoulli distribution
The Bernoulli distribution is the same as the binomial distribution
with parameter n = 1.
Poisson distribution
Parameter:
( > 0)
x
PF:
p(x) = e x! , x = 0, 1, 2, . . .
DF:
tables section.
PGF:
G(s) = exp ( (s 1))
MGF:
M (t) = exp ( (et 1))
Moments:
E[X] = , V ar(X) =
Coefficient
of skewness: 1
ACTL2131 & ACTL5101
c Katja Ignatieva
!
PGF:
G(s) =
h
ba+h
sb+h sa
h
< s 1
(b+h)t
2.2
=
at
e
e
h
MGF:
M (t) = ba+h
eht 1
1
1
(b a)(b a + 2h)
Moments:
E[X] = 2 (a + b), var(X) = 12
Continuous distributions
Standard Normal distribution - N(0,1)
Parameters: none
1 2
PF:
f (x) = 12 e 2 x , < x <
DF:
tables section.
1 2
MGF:
M (t) = e 2 t
Moments:
E[X] = 0, var(X) = 1
1 (1+r)
, r = 2, 3, 6, . . .
E[X r ] = 2r/2
(1+ r2 )
MGF:
Moments:
Exponential
Parameter:
PDF:
DF:
MGF:
Moments:
Coefficient
of skewness:
2 2
M (t) = et+ 2 t
E[X] = , var(X) = 2
distribution
( > 0)
f (x) = ex , x 0
F (x) = 1 ex
.1
M (t) = 1 t
,t<
E[X] = 1 , V ar(X) = 12
r = 1, 2, 3, . . .
E[X r ] = (1+r)
r
2
Uniform distribution (continuous) - U(a,b)

Parameters: a, b (a < b)
1
,a<x<b
PDF:
f (x) = ba
xa
DF:
F (x) = ba
- bt
.
1
1
at
MGF:
M (t) = (ba)
t e e
1
2
(b
Moments:
E[X] = 21 (a + b), V-ar(X) = 12
. a)
1
1
r
r+1
r+1
E[X ] = (ba) r+1 b
a
, r = 1, 2, 3, . . .
Beta distribution
Parameters: , ( > 0, > 0)
(+) 1
PDF:
f (x) = ()()
x
(1 x)1 , 0 < x < 1
Moments:
E[X] = +
, V ar(X) = (+)2
(++1)
E[X r ] =
Coefficient
of skewness:
2()
(++2)
(+)(+r)
()(++r) ,
++1
r = 1, 2, 3, . . .
Formulae and Tables
Page 1102 of 1113
Gamma distribution
Parameter:
, ( > 0, > 0)
PDF:
f (x) = ()
x1 ex , x 0
DF:
When 2 is an integer, probabilities for the gamma
distribution can be found using the relationship:
2X 22
.
,t<
MGF:
M (t) = 1 t
Moments:
E[X] = , V ar(X) = 2
E[X r ] = (+r)
()r r = 1, 2, 3, . . .
Coefficient
of skewness: 2
Chi-squared distribution - 2
The chi-squared distribution with degrees of freedom is the same as
the gamma distribution with parameters = 2 and = 12 .
The distribution function for the chi-squared distribution is tabulated in
the statistical tables section
Normal (Gaussian) distribution - N(, 2 )

Parameters: , 2 ( > 0)
<
.2 =
PF:
f (x) = 12 exp 21 x
, < x <
ACTL2131 & ACTL5101
c Katja Ignatieva
!
Uniform distribution (discrete)

Parameters: a, b, h (a < b, h > 0, b a is a multiple of h)
h
, <x = a, a +
PF:
p(x) = ba+h
= h, a + 2h, . . . , b h, b
PDF:
Moments:
%
<
=2 &
exp 21 log x
< 2
=
2
1 2
E[X] = e+ 2 , var(X) = e2+ e 1
f (x) =
1
1
2 x
E[X r ] = er+ 2 r
Moments:
, r = 1, 2, 3, . . .
Weibull distribution
Parameters: c, (c > 0, > 0)
PDF:
f (x) = cx1 ecx , < x <
DF:
F (x) = 1 <ecx =
Coefficient
< 2
=
of skewness:
e + 2
e 2 1
Pareto distribution (two parameter version)
Parameters: , , ( > 0, > 0)
,x>0
PDF:
f (x) = (+x)
<+1 =
DF:
F (x) = 1 +x
Moments:
2
1 ( > 1), var(X) = (1)2 (2)
r , r = 1, 2, 3, . . . , r <
E[X r ] = (r)(1+r)
()
E[X] =
Coefficient
of skewness:
2.3
2(+1)
(3)
k(k+1)2
k
1 ( > 1), var(X) = (1)2 (2)
r , r = 1, 2, 3, . . . , r <
E[X r ] = (r)(k+r)
()(k)
E[X] =
> 2)
( > 3)
1
Moments:
E[X r ] = 1 + r cr/
, r = 1, 2, 3, . . .
Burr distribution
Parameters: , , ( > 0, > 0, > 0)
x1
PDF:
f (x) = (+x
)+1 , x 0
=
<
DF:
F (x) = 1 +x
= <
= r/
<
Moments:
E[X r ] = r = r () r = 1, 2, 3, . . . , r <
Formulae and Tables
Page 1103 of 1113
Compound distributions
Conditional expectation and variance
E[Y ] = E[E[Y |X]]
V ar(Y ) = V ar(E[Y |X]) + E[V ar(Y |X)]
Moments of a compound distribution
If X1 , X2 , . . . are i.i.d random variables with mgf MX (t) and N is
an independent nonnegative integer-valued random variable, then
S = X1 + . . . + XN (with S = 0 when n = 0) has the following
properties:
Mean:
E[S] = E[N ] E[X]
2
Variance: V ar(S) = E[N ] V ar(X) + V ar(N ) (E[X])
MGF:
MS (t) = MN (log(MX (t)))
Compound Poisson distribution
Mean:
m1
Variance:
m2
Third central moment: m3
where = E[N ] and mr = E[X r ]
> 2)
ACTL2131 & ACTL5101
c Katja Ignatieva
!
Pareto distribution (three parameter version)

Parameters: , , k, ( > 0, > 0, k > 0)
(+k) xk1
PDF:
f (x) = ()(k)(+x)
+k , x > 0
Lognormal distribution
Parameters: , 2 ( > 0)
i=1
3.2
Parametric inference (normal model)

One sample
For a single sample of size n under the normal model X N (, 2 ):
(n1)S 2
X
tn1 and
2n1
2
S/ n
Two samples
For two independent samples of sizes m and n under the normal
2
models X N (X , X
) and Y N (Y , Y2 ):
2
2
SX
/X
Fm1,n1
S 2 /2
Y
3.3
2
Under the additional assumption that X
= Y2 :
(XY )(x Y )
1 1
tm+n2
Sp m
+n
.
1
2
2
where Sp = m+n2 (m 1)SX
+ (n 1)SY2 is the pooled sample
variance.
Maximum likelihood estimators
Asymptotic distribution
If 6 is the maximum likelihood estimator of parameter based on
a sample X, then 6 is asymptotically normally distributed with mean
and variance equal
N toI the Cramer-Rao
J lower bound
2
CRLB() = 1 E
2 log L(, X)
Likelihood ratio test%
&
maxL
H0
2(>p >p+q ) = 2 log max L 2q approximately (under H0 )
H0 H1
where
>p = max log L

H0
and
3.4
Linear regression model with normal errors

Model
Yi N ( + xi , 2 ), i = 1, 2, . . . , n
Intermediate calculations
n
n
8
8
sxx =
(xi x)2 =
x2i nx2
syy =
i=1
n
8
i=1
(yi y) =
n
8
yi2 ny 2
i=1
(xi x)(yi y) =
i=1
Page 1104 of 1113
Parameter estimates
6 6 = sxy
6 = y x,
sxx
n
1 8
2
6 = n2 (yi = y6i )2 =
i=1
Distribution of 6
#

tn2
2
# /sxx
n
8
xi yi nxy
i=1
1
n2
<
syy
s2xy
sx x
>p+q =
max log L
H0 H1
is the maximum log-likelihood for the

model under H0 (in which there are
p free parameters)
is the maximum log-likelihood for the
model under H0 H1 (in which there

are p + q free parameters)
Variance of predicted
mean =response
<
2
6 0 ) = 1 + (x0 x) 2
var(6
+ x
n
sxx
An additional 2 must be added to obtain the variance of the

predicted individual response.
Testing the correlation coefficient
sxy
r = xxx
syy
n2
If = 0, then r1r
tn2 .
2
Fisher Z <
transformation
=
1
zr N z , n3
approximately
<
=
1+r
where zr = tanh1 r = 12 log 1r
and z = tanh1 =
Sum of squares relationship
n
n
n
8
8
8
(yi y)2 =
(yi y6)2 +
(6
yi y)2
i=1
i=1
i=1
1
2
log
<
1+
1
Formulae and Tables
sxy =
i=1
n
8
ACTL2131 & ACTL5101
c Katja Ignatieva
!
3
Statistical Methods
3.1 Sample mean and variance
The random sample (x1 , x2 , . . . , xn ) has the following sample
moments:
n
8
x = n1 xi
Sample mean:
i=1 %
&
n
8
1
2
Sample variance: s = n1
x2i nx2
Analysis of Variance
Single factor normal model
Yij N ( + i , 2 ), i = 1, 2, . . . , k, j = 1, 2, . . . , ni
8
8
where n = ki=1 ni , with ki=1 ni i = 0
Single factor normal model
ni
ni
k 8
k 8
8
8
2
(yij y )2 =
Total:
SST =
yij
Between treaments: SSB =
i=1j=1
k
8
i=1
ni (y i y )2 =
i=1j=1
k 2
8
yi
i=1
ni
3.7
2
y
n
2
y
n
Residual:
SSR = SST SSB
Variance estimate
R
62 = SS
nk
Statistical test
Under the
N appropriate null hypothesis:
SSB
k1
SSR
nk
ACTL2131 & ACTL5101
c Katja Ignatieva
!
3.5
Bayesian methods
Relationship between posterior and prior distribution
Posterior Prior Likelihood
The posterior distribution f (|x) for the parameter is related to
the prior distributionf () via the likelihood function f (x|):
f (|x) f () f (x|)
Normal / normal model
If x is a random sample of size n from a N (, 2 ) distribution,
where 2 is known, and the prior distribution for the parameter is
N (0 , 02 ), then the posterior distribution for is:
|x N (<) , )2 ) =N <
=
N<
=
where ) =
nx
2
0
02
n
2
1
02
and )2 = 1
n
2
1
02
Fk1,nk
Statistical tables
Percentage Points for standard normal distribution

P
50%
45%
40%
35%
30%
x
0.0000
0.1257
0.2533
0.3853
0.5244
P
5.0%
5.0%
4.8%
4.6%
4.4%
x
1.6449
1.6646
1.6849
1.7060
1.7279
P
3.0%
2.9%
2.8%
2.7%
2.6%
x
1.8801
1.8957
1.9110
1.9268
1.9431
P
2.0%
1.9%
1.8%
1.7%
1.6%
x
2.0537
2.0749
2.0969
2.1201
2.1444
P
1.0%
0.9%
0.8%
0.7%
0.6%
x
2.3263
2.3656
2.4089
2.4573
2.5121
P
0.10%
0.09%
0.08%
0.07%
0.06%
x
3.0902
3.1214
3.1559
3.1947
3.2389
25%
20%
15%
10%
5%
0.6745
0.8416
1.0364
1.2816
1.6449
4.0%
3.8%
3.6%
3.4%
3.2%
1.7507
1.7744
1.7991
1.8250
1.8522
2.5%
2.4%
2.3%
2.2%
2.1%
1.9600
1.9774
1.9954
2.0141
2.0335
1.5%
1.4%
1.3%
1.2%
1.1%
2.1701
2.1973
2.2262
2.2571
2.2904
0.5%
0.4%
0.3%
0.2%
0.1%
2.5758
0.05%
2.6521
0.01%
2.7478
0.005%
2.8782
0.001%
3.0902 0.0005%
3.2905
3.7190
3.8906
4.2649
4.4172
Formulae and Tables
Page 1105 of 1113
Percentage Points for the standard normal distribution:

Formulae and Tables page 162
Cumulative density function for the standard normal distribution:
Formulae and Tables page 160-161
Percentage Points for the t distribution:
Formulae and Tables page 163
Probabilities for the 2 distribution:
Percentage Points for the 2 distribution:
Percentage Points for the F distribution:
Cumulative density function for the standard normal distribution
Cumulative density function for the standard normal distribution
x
0.40
0.41
0.42
0.43
0.44
(x)
0.65542
0.65910
0.66276
0.66640
0.67003
x
0.80
0.81
0.82
0.83
0.84
(x)
0.78814
0.79103
0.79389
0.79673
0.79955
x
1.20
1.21
1.22
1.23
1.24
(x)
0.88493
0.88686
0.88877
0.89065
0.89251
x
1.60
1.61
1.62
1.63
1.64
(x)
0.94520
0.94630
0.94738
0.94845
0.94950
x
2.00
2.01
2.02
2.03
2.04
(x)
0.97725
0.97778
0.97831
0.97882
0.97932
x
2.40
2.41
2.42
2.43
2.44
(x)
0.99180
0.99202
0.99224
0.99245
0.99266
x
2.80
2.81
2.82
2.83
2.84
(x)
0.99744
0.99752
0.99760
0.99767
0.99774
x
3.20
3.21
3.22
3.23
3.24
(x)
0.99931
0.99934
0.99936
0.99938
0.99940
x
3.60
3.61
3.62
3.63
3.64
(x)
0.99984
0.99985
0.99985
0.99986
0.99986
x
4.00
4.01
4.02
4.03
4.04
(x)
0.99997
0.99997
0.99997
0.99997
0.99997
0.05
0.06
0.07
0.08
0.09
0.51994
0.52392
0.52790
0.53188
0.53586
0.45
0.46
0.47
0.48
0.49
0.67364
0.67724
0.68082
0.68439
0.68793
0.85
0.86
0.87
0.88
0.89
0.80234
0.80511
0.80785
0.81057
0.81327
1.25
1.26
1.27
1.28
1.29
0.89435
0.89617
0.89796
0.89973
0.90147
1.65
1.66
1.67
1.68
1.69
0.95053
0.95154
0.95254
0.95352
0.95449
2.05
2.06
2.07
2.08
2.09
0.97982
0.98030
0.98077
0.98124
0.98169
2.45
2.46
2.47
2.48
2.49
0.99286
0.99305
0.99324
0.99343
0.99361
2.85
2.86
2.87
2.88
2.89
0.99781
0.99788
0.99795
0.99801
0.99807
3.25
3.26
3.27
3.28
3.29
0.99942
0.99944
0.99946
0.99948
0.99950
3.65
3.66
3.67
3.68
3.69
0.99987
0.99987
0.99988
0.99988
0.99989
4.05
4.06
4.07
4.08
4.09
0.99997
0.99998
0.99998
0.99998
0.99998
0.10
0.11
0.12
0.13
0.14
0.53983
0.54380
0.54776
0.55172
0.55567
0.50
0.51
0.52
0.53
0.54
0.69146
0.69497
0.69847
0.70194
0.70540
0.90
0.91
0.92
0.93
0.94
0.81594
0.81859
0.82121
0.82381
0.82639
1.30
1.31
1.32
1.33
1.34
0.90320
0.90490
0.90658
0.90824
0.90988
1.70
1.71
1.72
1.73
1.74
0.95543
0.95637
0.95728
0.95818
0.95907
2.10
2.11
2.12
2.13
2.14
0.98214
0.98257
0.98300
0.98341
0.98382
2.50
2.51
2.52
2.53
2.54
0.99379
0.99396
0.99413
0.99430
0.99446
2.90
2.91
2.92
2.93
2.94
0.99813
0.99819
0.99825
0.99831
0.99836
3.30
3.31
3.32
3.33
3.34
0.99952
0.99953
0.99955
0.99957
0.99958
3.70
3.71
3.72
3.73
3.74
0.99989
0.99990
0.99990
0.99990
0.99991
4.10
4.11
4.12
4.13
4.14
0.99998
0.99998
0.99998
0.99998
0.99998
0.15
0.16
0.17
0.18
0.19
0.55962
0.56356
0.56749
0.57142
0.57535
0.55
0.56
0.57
0.58
0.59
0.70884
0.71226
0.71566
0.71904
0.72240
0.95
0.96
0.97
0.98
0.99
0.82894
0.83147
0.83398
0.83646
0.83891
1.35
1.36
1.37
1.38
1.39
0.91149
0.91309
0.91466
0.91621
0.91774
1.75
1.76
1.77
1.78
1.79
0.95994
0.96080
0.96164
0.96246
0.96327
2.15
2.16
2.17
2.18
2.19
0.98422
0.98461
0.98500
0.98537
0.98574
2.55
2.56
2.57
2.58
2.59
0.99461
0.99477
0.99492
0.99506
0.99520
2.95
2.96
2.97
2.98
2.99
0.99841
0.99846
0.99851
0.99856
0.99861
3.35
3.36
3.37
3.38
3.39
0.99960
0.99961
0.99962
0.99964
0.99965
3.75
3.76
3.77
3.78
3.79
0.99991
0.99992
0.99992
0.99992
0.99992
4.15
4.16
4.17
4.18
4.19
0.99998
0.99998
0.99998
0.99999
0.99999
0.20
0.21
0.22
0.23
0.24
0.57926
0.58317
0.58706
0.59095
0.59483
0.60
0.61
0.62
0.63
0.64
0.72575
0.72907
0.73237
0.73565
0.73891
1.00
1.01
1.02
1.03
1.04
0.84134
0.84375
0.84614
0.84849
0.85083
1.40
1.41
1.42
1.43
1.44
0.91924
0.92073
0.92220
0.92364
0.92507
1.80
1.81
1.82
1.83
1.84
0.96407
0.96485
0.96562
0.96638
0.96712
2.20
2.21
2.22
2.23
2.24
0.98610
0.98645
0.98679
0.98713
0.98745
2.60
2.61
2.62
2.63
2.64
0.99534
0.99547
0.99560
0.99573
0.99585
3.00
3.01
3.02
3.03
3.04
0.99865
0.99869
0.99874
0.99878
0.99882
3.40
3.41
3.42
3.43
3.44
0.99966
0.99968
0.99969
0.99970
0.99971
3.80
3.81
3.82
3.83
3.84
0.99993
0.99993
0.99993
0.99994
0.99994
4.20
4.21
4.22
4.23
4.24
0.99999
0.99999
0.99999
0.99999
0.99999
0.25
0.26
0.27
0.28
0.29
0.59871
0.60257
0.60642
0.61026
0.61409
0.65
0.66
0.67
0.68
0.69
0.74215
0.74537
0.74857
0.75175
0.75490
1.05
1.06
1.07
1.08
1.09
0.85314
0.85543
0.85769
0.85993
0.86214
1.45
1.46
1.47
1.48
1.49
0.92647
0.92785
0.92922
0.93056
0.93189
1.85
1.86
1.87
1.88
1.89
0.96784
0.96856
0.96926
0.96995
0.97062
2.25
2.26
2.27
2.28
2.29
0.98778
0.98809
0.98840
0.98870
0.98899
2.65
2.66
2.67
2.68
2.69
0.99598
0.99609
0.99621
0.99632
0.99643
3.05
3.06
3.07
3.08
3.09
0.99886
0.99889
0.99893
0.99896
0.99900
3.45
3.46
3.47
3.48
3.49
0.99972
0.99973
0.99974
0.99975
0.99976
3.85
3.86
3.87
3.88
3.89
0.99994
0.99994
0.99995
0.99995
0.99995
4.25
4.26
4.27
4.28
4.29
0.99999
0.99999
0.99999
0.99999
0.99999
0.30
0.31
0.32
0.33
0.34
0.61791
0.62172
0.62552
0.62930
0.63307
0.70
0.71
0.72
0.73
0.74
0.75804
0.76115
0.76424
0.76730
0.77035
1.10
1.11
1.12
1.13
1.14
0.86433
0.86650
0.86864
0.87076
0.87286
1.50
1.51
1.52
1.53
1.54
0.93319
0.93448
0.93574
0.93699
0.93822
1.90
1.91
1.92
1.93
1.94
0.97128
0.97193
0.97257
0.97320
0.97381
2.30
2.31
2.32
2.33
2.34
0.98928
0.98956
0.98983
0.99010
0.99036
2.70
2.71
2.72
2.73
2.74
0.99653
0.99664
0.99674
0.99683
0.99693
3.10
3.11
3.12
3.13
3.14
0.99903
0.99906
0.99910
0.99913
0.99916
3.50
3.51
3.52
3.53
3.54
0.99977
0.99978
0.99978
0.99979
0.99980
3.90
3.91
3.92
3.93
3.94
0.99995
0.99995
0.99996
0.99996
0.99996
4.30
4.31
4.32
4.33
4.34
0.99999
0.99999
0.99999
0.99999
0.99999
0.35
0.36
0.37
0.38
0.39
0.63683
0.64058
0.64431
0.64803
0.65173
0.75
0.76
0.77
0.78
0.79
0.77337
0.77637
0.77935
0.78230
0.78524
1.15
1.16
1.17
1.18
1.19
0.87493
0.87698
0.87900
0.88100
0.88298
1.55
1.56
1.57
1.58
1.59
0.93943
0.94062
0.94179
0.94295
0.94408
1.95
1.96
1.97
1.98
1.99
0.97441
0.97500
0.97558
0.97615
0.97670
2.35
2.36
2.37
2.38
2.39
0.99061
0.99086
0.99111
0.99134
0.99158
2.75
2.76
2.77
2.78
2.79
0.99702
0.99711
0.99720
0.99728
0.99736
3.15
3.16
3.17
3.18
3.19
0.99918
0.99921
0.99924
0.99926
0.99929
3.55
3.56
3.57
3.58
3.59
0.99981
0.99981
0.99982
0.99983
0.99983
3.95
3.96
3.97
3.98
3.99
0.99996
0.99996
0.99996
0.99997
0.99997
4.35
4.36
4.37
4.38
4.39
0.99999
0.99999
0.99999
0.99999
0.99999
0.40
0.65542
0.80
0.78814
1.20
0.88493
1.60
0.94520
2.00
0.97725
2.40
0.99180
2.80
0.99744
3.20
0.99931
3.60
0.99984
4.00
0.99997
4.40
0.99999
Formulae and Tables
Page 1106 of 1113
(x)
0.50000
0.50399
0.50798
0.51197
0.51595
ACTL2131 & ACTL5101
c Katja Ignatieva
!
x
0.00
0.01
0.02
0.03
0.04
Percentage Points for the t distribution
Probabilities for the 2 distribution
30%
25%
20%
15% 10% 5%
2.5% 1%
0.5% 0.1% 0.05%
0.3249
0.2887
0.2767
0.2707
0.7265
0.6172
0.5844
0.5686
1.0000
0.8165
0.7649
0.7407
1.376
1.061
0.9785
0.9410
1.963
1.386
1.250
1.190
3.078
1.866
1.638
1.533
6.314
2.920
2.353
2.132
12.71
4.303
3.182
2.776
31.82
6.965
4.541
3.747
63.66
9.925
5.841
4.604
318.3
22.33
10.21
7.173
636.6
31.60
12.92
8.610
5
6
7
8
9
0.2672
0.2648
0.2632
0.2619
0.2610
0.5594
0.5534
0.5491
0.5459
0.5435
0.7267
0.7176
0.7111
0.7064
0.7072
0.9195
0.9057
0.8960
0.8889
0.8834
1.156
1.134
1.119
1.108
1.100
1.476
1.440
1.415
1.397
1.383
2.015
1.943
1.895
1.860
1.833
2.571
2.447
2.365
2.306
2.262
3.365
3.143
2.998
2.896
2.821
4.032
3.707
3.499
3.355
3.250
5.894
5.208
4.785
4.501
4.297
6.869
5.959
5.408
5.041
4.781
10
11
12
13
14
0.2602
0.2596
0.2590
0.2586
0.2582
0.5415
0.5399
0.5386
0.5375
0.5366
0.6998
0.6974
0.6955
0.6938
0.6924
0.8791
0.8755
0.8726
0.8702
0.8681
1.093
1.088
1.083
1.079
1.076
1.372
1.363
1.356
1.350
1.345
1.812
1.796
1.782
1.771
1.761
2.228
2.201
2.176
2.160
2.145
2.764
2.718
2.681
2.650
2.624
3.169
3.106
3.055
3.012
2.977
4.144
4.025
3.930
3.852
3.787
4.587
4.437
4.318
4.221
4.140
15
16
17
18
19
0.2579
0.2576
0.2573
0.2571
0.2569
0.5357
0.5350
0.5344
0.5338
0.5333
0.6912
0.6901
0.6892
0.6884
0.6876
0.8662
0.8647
0.8633
0.8620
0.8610
1.074
1.071
1.069
1.067
1.066
1.341
1.337
1.333
1.330
1.328
1.753
1.746
1.740
1.734
1.729
2.131
2.120
2.110
2.101
2.093
2.602
2.583
2.567
2.552
2.539
2.947
2.921
2.898
2.878
2.861
3.733
3.686
3.646
3.610
3.579
20
21
22
23
24
0.2567
0.2566
0.2564
0.2563
0.2562
0.5329
0.5325
0.5321
0.5317
0.5314
0.6870
0.6864
0.6858
0.6853
0.6848
0.8600
0.8591
0.8583
0.8575
0.8569
1.064
1.063
1.061
1.060
1.059
1.325
1.323
1.321
1.319
1.318
1.725
1.721
1.717
1.714
1.711
2.086
2.080
2.074
2.069
2.064
2.528
2.518
2.508
2.500
2.492
2.845
2.831
2.819
2.807
2.797
25
26
27
28
29
0.2561
0.2560
0.2559
0.2558
0.2557
0.5312
0.5309
0.5306
0.5304
0.5302
0.6844
0.6840
0.6837
0.6834
0.6830
0.8562
0.8557
0.8551
0.8546
0.8542
1.058
1.058
1.057
1.056
1.055
1.316
1.315
1.314
1.313
1.311
1.708
1.706
1.703
1.701
1.699
2.060
2.056
2.052
2.048
2.045
2.485
2.479
2.473
2.467
2.462
30
32
34
36
38
0.2556
0.2555
0.2553
0.2552
0.2551
0.5300
0.5297
0.5294
0.5291
0.5288
0.6828
0.6822
0.6818
0.6814
0.6810
0.8538
0.8530
0.8523
0.8517
0.8512
1.055
1.054
1.052
1.052
1.051
1.310
1.309
1.307
1.306
1.304
1.697
1.694
1.691
1.688
1.686
2.042
2.037
2.032
2.028
2.024
40
50
60
120
0.2550
0.2547
0.2545
0.2539
0.2533
0.5286
0.5278
0.5272
0.5258
0.5244
0.6807
0.6794
0.6786
0.6765
0.6745
0.8507
0.8489
0.8477
0.8446
0.8416
1.050
1.047
1.045
1.041
1.036
1.303
1.299
1.296
1.289
1.282
1.684
1.676
1.671
1.658
1.645
2.021
2.009
2.000
1.980
1.960
=
x
0.0
0.1
0.2
0.3
0.4
0.0000
0.2482
0.3453
0.4161
0.4729
x
4.0
4.1
4.2
4.3
4.4
0.9545
0.9571
0.9596
0.9619
0.9641
x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0.5205
0.5614
0.5972
0.6289
0.6572
4.5
4.6
4.7
4.8
4.9
0.9661
0.9680
0.9698
0.9715
0.9731
1.0
1.1
1.2
1.3
1.4
0.6827
0.7057
0.7267
0.7458
0.7633
5.0
5.1
5.2
5.3
5.4
4.073
4.015
3.965
3.922
3.883
1.5
1.6
1.7
1.8
1.9
0.7793
0.7941
0.8077
0.8203
0.8319
3.522
3.527
3.505
3.485
3.467
3.850
3.819
3.792
3.768
3.745
2.0
2.1
2.2
2.3
2.4
2.787
2.779
2.771
2.763
2.756
3.450
3.435
3.421
3.408
3.396
3.725
3.707
3.689
3.674
3.660
2.457
2.449
2.441
2.434
2.429
2.750
2.738
2.728
2.719
2.712
3.385
3.365
3.348
3.333
3.319
3.646
3.622
3.601
3.582
3.566
2.423
2.403
2.390
2.358
2.326
2.704
2.678
2.660
2.617
2.576
3.307
3.261
3.232
3.160
3.090
3.551
3.496
3.460
3.373
3.291
0.0000
0.0488
0.0952
0.1393
0.1813
x
4.0
4.1
4.2
4.3
4.4
0.5
0.6
0.7
0.8
0.9
0.2212
0.2592
0.2953
0.3297
0.3624
0.9747
0.9761
0.9774
0.9787
0.9799
1.0
1.1
1.2
1.3
1.4
5.5
5.6
5.7
5.8
5.9
0.9810
0.9820
0.9830
0.9840
0.9849
0.8427
0.8527
0.8620
0.8706
0.8787
6.0
6.1
6.2
6.3
6.4
2.5
2.6
2.7
2.8
2.9
0.8862
0.8931
0.8997
0.9057
0.9114
3.0
3.1
3.2
3.3
3.4
0.8647
0.8713
0.8775
0.8835
0.8892
x
0.0
0.1
0.2
0.3
0.4
4.5
4.6
4.7
4.8
4.9
0.8946
0.8997
0.9046
0.9093
0.9137
0.3935
0.4231
0.4512
0.4780
0.5034
5.0
5.1
5.2
5.3
5.4
1.5
1.6
1.7
1.8
1.9
0.5276
0.5507
0.5726
0.5934
0.6133
0.9857
0.9865
0.9872
0.9879
0.9886
2.0
2.1
2.2
2.3
2.4
6.5
6.6
6.7
6.8
6.9
0.9892
0.9898
0.9904
0.9909
0.9914
0.9167
0.9217
0.9264
0.9307
0.9348
7.0
7.1
7.2
7.3
7.4
3.5
3.6
3.7
3.8
3.9
0.9386
0.9422
0.9456
0.9487
0.9517
4.0
0.9545
0.0000
0.0082
0.0224
0.0400
0.0598
x
4.0
4.1
4.2
4.3
4.4
0.7385
0.7491
0.7593
0.7692
0.7786
0.5
0.6
0.7
0.8
0.9
0.0811
0.1036
0.1268
0.1505
0.1746
4.5
4.6
4.7
4.8
4.9
0.7877
0.7965
0.8049
0.8130
0.8207
0.9179
0.9219
0.9257
0.9293
0.9328
1.0
1.1
1.2
1.3
1.4
0.1987
0.2229
0.2470
0.2709
0.2945
5.0
5.1
5.2
5.3
5.4
0.8282
0.8354
0.8423
0.8489
0.8553
5.5
5.6
5.7
5.8
5.9
0.9361
0.9392
0.9422
0.9450
0.9477
1.5
1.6
1.7
1.8
1.9
0.3177
0.3406
0.3631
0.3851
0.4066
5.5
5.6
5.7
5.8
5.9
0.8614
0.8672
0.8728
0.8782
0.8834
0.6321
0.6501
0.6671
0.6834
0.6988
6.0
6.1
6.2
6.3
6.4
0.9502
0.9526
0.9550
0.9571
0.9592
2.0
2.1
2.2
2.3
2.4
0.4276
0.4481
0.4681
0.4875
0.5064
6.0
6.1
6.2
6.3
6.4
0.8884
0.8932
0.8977
0.9021
0.9063
2.5
2.6
2.7
2.8
2.9
0.7135
0.7275
0.7408
0.7534
0.7654
6.5
6.6
6.7
6.8
6.9
0.9612
0.9631
0.9649
0.9666
0.9683
2.5
2.6
2.7
2.8
2.9
0.5247
0.5425
0.5598
0.5765
0.5927
6.5
6.6
6.7
6.8
6.9
0.9103
0.9142
0.9179
0.9214
0.9248
0.9918
0.9923
0.9927
0.9931
0.9935
3.0
3.1
3.2
3.3
3.4
0.7769
0.7878
0.7981
0.8080
0.8173
7.0
7.1
7.2
7.3
7.4
0.9698
0.9713
0.9727
0.9740
0.9753
3.0
3.1
3.2
3.3
3.4
0.6084
0.6235
0.6382
0.6524
0.6660
7.0
7.1
7.2
7.3
7.4
0.9281
0.9312
0.9342
0.9371
0.9398
7.5
7.6
7.7
7.8
7.9
0.9938
0.9942
0.9945
0.9948
0.9951
3.5
3.6
3.7
3.8
3.9
0.8262
0.8347
0.8428
0.8504
0.8577
7.5
7.6
7.7
7.8
7.9
0.9765
0.9776
0.9787
0.9798
0.9807
3.5
3.6
3.7
3.8
3.9
0.6792
0.6920
0.7043
0.7161
0.7275
7.5
7.6
7.7
7.8
7.9
0.9424
0.9450
0.9474
0.9497
0.9519
8.0
0.9953
4.0
0.8647
8.0
0.9817
4.0
0.7385
8.0
0.9540
Formulae and Tables
Page 1107 of 1113
40%
ACTL2131 & ACTL5101
c Katja Ignatieva
!
P=
1
2
3
4
10
11
12
13
14
0.0265
0.0902
0.1734
0.2642
0.3554
0.4422
0.5221
0.5940
0.6575
0.7127
0.7603
0.8009
0.8352
0.8641
0.8883
0.9084
0.9251
0.9389
0.9503
0.9596
0.9672
0.9734
0.9785
0.9826
0.9860
0.9887
0.9909
0.9927
0.9941
0.9953
0.9962
0.9970
0.9976
0.9981
0.9985
0.9988
0.9990
0.9992
0.9994
0.9995
0.9997
0.9998
0.9999
0.9999
0.9999
1.0000
1.0000
1.0000
1.0000
1.0000
0.0079
0.0374
0.0869
0.1509
0.2235
0.3000
0.3766
0.4506
0.5201
0.5841
0.6421
0.6938
0.7394
0.7794
0.8140
0.8438
0.8693
0.8909
0.9093
0.9248
0.9378
0.9486
0.9577
0.9652
0.9715
0.9766
0.9809
0.9844
0.9873
0.9896
0.9916
0.9932
0.9944
0.9955
0.9964
0.9971
0.9976
0.9981
0.9984
0.9988
0.9992
0.9995
0.9997
0.9998
0.9999
0.9999
0.9999
1.0000
1.0000
1.0000
0.0022
0.0144
0.0405
0.0803
0.1315
0.1912
0.2560
0.3233
0.3907
0.4562
0.5185
0.5768
0.6304
0.6792
0.7229
0.7619
0.7963
0.8264
0.8527
0.8753
0.8949
0.9116
0.9259
0.9380
0.9483
0.9570
0.9643
0.9704
0.9755
0.9797
0.9833
0.9862
0.9887
0.9907
0.9924
0.9938
0.9949
0.9958
0.9966
0.9972
0.9982
0.9988
0.9992
0.9995
0.9997
0.9998
0.9999
0.9999
0.9999
1.0000
0.0006
0.0052
0.0177
0.0402
0.0729
0.1150
0.1648
0.2202
0.2793
0.3400
0.4008
0.4603
0.5173
0.5711
0.6213
0.6674
0.7094
0.7473
0.7813
0.8114
0.8380
0.8614
0.8818
0.8994
0.9147
0.9279
0.9392
0.9488
0.9570
0.9640
0.9699
0.9749
0.9791
0.9826
0.9856
0.9880
0.9901
0.9918
0.9932
0.9944
0.9962
0.9975
0.9983
0.9989
0.9992
0.9995
0.9997
0.9998
0.9999
0.9999
0.0001
0.0018
0.0073
0.0190
0.0383
0.0656
0.1008
0.1429
0.1906
0.2424
0.2970
0.3528
0.4086
0.4634
0.5162
0.5665
0.6138
0.6577
0.6981
0.7350
0.7683
0.7983
0.8251
0.8488
0.8697
0.8882
0.9042
0.9182
0.9304
0.9409
0.9499
0.9576
0.9642
0.9699
0.9747
0.9788
0.9822
0.9851
0.9876
0.9897
0.9929
0.9951
0.9966
0.9977
0.9984
0.9989
0.9993
0.9995
0.9997
0.9998
0.0000
0.0006
0.0029
0.0085
0.0191
0.0357
0.0589
0.0886
0.1245
0.1657
0.2113
0.2601
0.3110
0.3629
0.4148
0.4659
0.5154
0.5627
0.6075
0.6495
0.6885
0.7243
0.7570
0.7867
0.8134
0.8374
0.8587
0.8777
0.8944
0.9091
0.9219
0.9331
0.9429
0.9513
0.9586
0.9648
0.9702
0.9748
0.9787
0.9821
0.9873
0.9911
0.9938
0.9957
0.9970
0.9980
0.9986
0.9990
0.9994
0.9996
0.0000
0.0002
0.0011
0.0037
0.0091
0.0186
0.0329
0.0527
0.0780
0.1088
0.1446
0.1847
0.2283
0.2746
0.3225
0.3712
0.4199
0.4679
0.5146
0.5595
0.6022
0.6425
0.6801
0.7149
0.7470
0.7763
0.8030
0.8270
0.8486
0.8679
0.8851
0.9004
0.9138
0.9256
0.9360
0.9450
0.9529
0.9597
0.9656
0.9707
0.9789
0.9849
0.9893
0.9924
0.9947
0.9963
0.9974
0.9982
0.9988
0.9991
0.0000
0.0001
0.0004
0.0015
0.0042
0.0093
0.0177
0.0301
0.0471
0.0688
0.0954
0.1266
0.1620
0.2009
0.2427
0.2867
0.3321
0.3781
0.4242
0.4696
0.5140
0.5567
0.5976
0.6364
0.6727
0.7067
0.7381
0.7670
0.7935
0.8175
0.8393
0.8589
0.8764
0.8921
0.9061
0.9184
0.9293
0.9389
0.9473
0.9547
0.9666
0.9756
0.9823
0.9873
0.9909
0.9935
0.9954
0.9968
0.9977
0.9984
0.0000
0.0000
0.0001
0.0006
0.0018
0.0045
0.0091
0.0166
0.0274
0.0420
0.0608
0.0839
0.1112
0.1424
0.1771
0.2149
0.2551
0.2971
0.3403
0.3840
0.4278
0.4711
0.5134
0.5543
0.5936
0.6310
0.6662
0.6993
0.7301
0.7586
0.7848
0.8088
0.8306
0.8504
0.8683
0.8843
0.8987
0.9115
0.9228
0.9329
0.9496
0.9625
0.9723
0.9797
0.9852
0.9893
0.9923
0.9945
0.9961
0.9972
0.0000
0.0000
0.0000
0.0002
0.0008
0.0021
0.0046
0.0088
0.0154
0.0248
0.0375
0.0538
0.0739
0.0978
0.1254
0.1564
0.1904
0.2271
0.2658
0.3061
0.3474
0.3892
0.4310
0.4724
0.5129
0.5522
0.5900
0.6262
0.6604
0.6926
0.7228
0.7509
0.7768
0.8007
0.8226
0.8425
0.8606
0.8769
0.8916
0.9048
0.9271
0.9446
0.9583
0.9689
0.9769
0.9830
0.9876
0.9910
0.9935
0.9953
0.0000
0.0000
0.0000
0.0001
0.0003
0.0009
0.0022
0.0045
0.0084
0.0142
0.0224
0.0335
0.0477
0.0653
0.0863
0.1107
0.1383
0.1689
0.2022
0.2378
0.2752
0.3140
0.3536
0.3937
0.4338
0.4735
0.5124
0.5503
0.5868
0.6218
0.6551
0.6866
0.7162
0.7438
0.7695
0.7932
0.8151
0.8351
0.8533
0.8699
0.8984
0.9214
0.9397
0.9542
0.9654
0.9741
0.9807
0.9858
0.9895
0.9924
Probabilities for the 2 distribution

=
x
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
15
16
17
18
19
20
21
22
23
24
25
0.0004
0.0023
0.0079
0.0203
0.0424
0.0762
0.1225
0.1803
0.2474
0.3210
0.3977
0.4745
0.5486
0.6179
0.6811
0.7373
0.7863
0.8281
0.8632
0.8922
0.9159
0.9349
0.9501
0.9620
0.9713
0.9784
0.9839
0.9881
0.9912
0.9936
0.9953
0.9966
0.9975
0.9982
0.9987
0.9991
0.9994
0.9995
0.9997
0.9998
0.9998
0.9999
0.9999
0.9999
1.0000
1.0000
1.0000
1.0000
0.0002
0.0011
0.0042
0.0119
0.0267
0.0511
0.0866
0.1334
0.1905
0.2560
0.3272
0.4013
0.4754
0.5470
0.6144
0.6761
0.7313
0.7798
0.8215
0.8568
0.8863
0.9105
0.9302
0.9460
0.9585
0.9684
0.9761
0.9820
0.9865
0.9900
0.9926
0.9946
0.9960
0.9971
0.9979
0.9985
0.9989
0.9992
0.9994
0.9996
0.9997
0.9998
0.9999
0.9999
0.9999
1.0000
1.0000
1.0000
0.0001
0.0005
0.0022
0.0068
0.0165
0.0335
0.0597
0.0964
0.1434
0.1999
0.2638
0.3329
0.4045
0.4762
0.5456
0.6112
0.6715
0.7258
0.7737
0.8153
0.8507
0.8806
0.9053
0.9255
0.9419
0.9551
0.9655
0.9737
0.9800
0.9850
0.9887
0.9916
0.9938
0.9954
0.9966
0.9975
0.9982
0.9987
0.9991
0.9993
0.9995
0.9997
0.9998
0.9998
0.9999
0.9999
0.9999
1.0000
0.0000
0.0002
0.0011
0.0038
0.0099
0.0214
0.0403
0.0681
0.1056
0.1528
0.2084
0.2709
0.3380
0.4075
0.4769
0.5443
0.6082
0.6672
0.7206
0.7680
0.8094
0.8450
0.8751
0.9002
0.9210
0.9379
0.9516
0.9626
0.9712
0.9780
0.9833
0.9874
0.9905
0.9929
0.9948
0.9961
0.9972
0.9979
0.9985
0.9989
0.9992
0.9994
0.9996
0.9997
0.9998
0.9998
0.9999
0.9999
0.0000
0.0001
0.0006
0.0021
0.0058
0.0133
0.0265
0.0471
0.0762
0.1144
0.1614
0.2163
0.2774
0.3427
0.4101
0.4776
0.5432
0.6054
0.6632
0.7157
0.7627
0.8038
0.8395
0.8698
0.8953
0.9166
0.9340
0.9482
0.9596
0.9687
0.9760
0.9816
0.9860
0.9894
0.9921
0.9941
0.9956
0.9967
0.9976
0.9982
0.9987
0.9991
0.9993
0.9995
0.9996
0.9997
0.9998
0.9999
0.0000
0.0000
0.0003
0.0011
0.0033
0.0081
0.0171
0.0318
0.0538
0.0839
0.1226
0.1695
0.2236
0.2834
0.3470
0.4126
0.4782
0.5421
0.6029
0.6595
0.7112
0.7576
0.7986
0.8342
0.8647
0.8906
0.9122
0.9301
0.9448
0.9567
0.9663
0.9739
0.9799
0.9846
0.9883
0.9911
0.9933
0.9950
0.9963
0.9972
0.9980
0.9985
0.9989
0.9992
0.9994
0.9996
0.9997
0.9998
0.0000
0.0000
0.0001
0.0006
0.0019
0.0049
0.0108
0.0211
0.0372
0.0604
0.0914
0.1304
0.1770
0.2303
0.2889
0.3510
0.4149
0.4787
0.5411
0.6005
0.6560
0.7069
0.7528
0.7936
0.8291
0.8598
0.8860
0.9080
0.9263
0.9414
0.9538
0.9638
0.9718
0.9781
0.9832
0.9871
0.9902
0.9926
0.9944
0.9958
0.9969
0.9977
0.9983
0.9987
0.9991
0.9993
0.9995
0.9996
0.0000
0.0000
0.0001
0.0003
0.0010
0.0028
0.0067
0.0137
0.0253
0.0426
0.0668
0.0985
0.1378
0.1841
0.2366
0.2940
0.3547
0.4170
0.4793
0.5401
0.5983
0.6528
0.7029
0.7483
0.7888
0.8243
0.8551
0.8815
0.9039
0.9226
0.9381
0.9509
0.9613
0.9696
0.9763
0.9817
0.9859
0.9892
0.9918
0.9937
0.9953
0.9965
0.9973
0.9980
0.9985
0.9989
0.9992
0.9994
0.0000
0.0000
0.0000
0.0001
0.0005
0.0016
0.0040
0.0087
0.0168
0.0295
0.0480
0.0731
0.1054
0.1447
0.1907
0.2425
0.2988
0.3581
0.4189
0.4797
0.5392
0.5962
0.6497
0.6991
0.7440
0.7842
0.8197
0.8506
0.8772
0.8999
0.9189
0.9348
0.9480
0.9587
0.9675
0.9745
0.9802
0.9846
0.9882
0.9909
0.9931
0.9947
0.9960
0.9970
0.9978
0.9983
0.9988
0.9991
0.0000
0.0000
0.0000
0.0001
0.0003
0.0009
0.0024
0.0055
0.0110
0.0201
0.0339
0.0533
0.0792
0.1119
0.1513
0.1970
0.2480
0.3032
0.3613
0.4207
0.4802
0.5384
0.5942
0.6468
0.6955
0.7400
0.7799
0.8152
0.8462
0.8730
0.8959
0.9153
0.9316
0.9451
0.9562
0.9653
0.9727
0.9786
0.9833
0.9871
0.9901
0.9924
0.9942
0.9956
0.9967
0.9975
0.9981
0.9986
0.0000
0.0000
0.0000
0.0000
0.0001
0.0005
0.0014
0.0033
0.0071
0.0134
0.0235
0.0383
0.0586
0.0852
0.1182
0.1576
0.2029
0.2532
0.3074
0.3643
0.4224
0.4806
0.5376
0.5924
0.6441
0.6921
0.7361
0.7757
0.8110
0.8420
0.8689
0.8921
0.9118
0.9284
0.9423
0.9537
0.9632
0.9708
0.9770
0.9820
0.9860
0.9892
0.9916
0.9936
0.9951
0.9963
0.9972
0.9979
Formulae and Tables
Page 1108 of 1113
ACTL2131 & ACTL5101
c Katja Ignatieva
!
=
x
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
10.5
11.0
11.5
12.0
12.5
13.0
13.5
14.0
14.5
15.0
15.5
16.0
16.5
17.0
17.5
18.0
18.5
19.0
19.5
20
21
22
23
24
25
26
27
28
29
30
Percentage points for the 2 distribution
Percentage points for the 2 distribution

99.95%
99.9%
99.5%
99%
97.5%
95%
90%
3.927E-07 1.571E-06 3.927E-05 1.571E-04 9.821E-04 0.003932 0.01579

0.001000 0.00200
0.01003
0.02010
0.05064
0.1026 0.2107
0.01528
0.02430
0.07172
0.1148
0.2158
0.3518 0.5844
0.06392
0.09080
0.2070
0.2971
0.4844
0.7107 1.064
80%
70%
0.06418
0.4463
1.005
1.649
0.1485
0.7133
1.424
2.195
60% P=
0.2750 1
1.022 2
1.869 3
2.753 4
50%
40%
30%
20%
10%
5%
2.5%
1%
0.5%
0.1%
0.05%
0.4549
1.386
2.366
3.357
0.7083
1.833
2.946
4.045
1.074
2.408
3.665
4.878
1.642
3.219
4.642
5.989
2.706
4.605
6.251
7.779
3.841
5.991
7.815
9.488
5.024
7.378
9.348
11.14
6.635
9.210
11.34
13.28
7.879
10.60
12.84
14.86
10.83
13.82
16.27
18.47
12.12
15.20
17.73
20.00
0.1581
0.2994
0.4849
0.7104
0.9717
0.2102
0.3811
0.5985
0.8571
1.152
0.4117
0.6757
0.9893
1.344
1.735
0.5543
0.8721
1.239
1.646
2.088
0.8312
1.237
1.690
2.180
2.700
1.145
1.635
2.167
2.733
3.325
1.610
2.204
2.833
3.490
4.168
2.343
3.070
3.822
4.594
5.380
3.000
3.828
4.671
5.527
6.393
3.655
4.570
5.493
6.423
7.357
5
6
7
8
9
4.351
5.348
6.346
7.344
8.343
5.132
6.211
7.283
8.351
9.414
6.064
7.231
8.383
9.524
10.66
7.289
8.558
9.803
11.03
12.24
9.236
10.64
12.02
13.36
14.68
11.07
12.59
14.07
15.51
16.92
12.83
14.45
16.01
17.53
19.02
15.09
16.81
18.48
20.09
21.67
16.75
18.55
20.28
21.95
23.59
20.52
22.46
24.32
26.12
27.88
22.11
24.10
26.02
27.87
29.67
10
11
12
13
14
1.265
1.587
1.934
2.305
2.697
1.479
1.834
2.214
2.617
3.041
2.156
2.603
3.074
3.565
4.075
2.558
3.053
3.571
4.107
4.660
3.247
3.816
4.404
5.009
5.629
3.940
4.575
5.226
5.892
6.571
4.865
5.578
6.304
7.042
7.790
6.179
6.989
7.807
8.634
9.467
7.267
8.148
9.034
9.926
10.82
8.295
9.237
10.18
11.13
12.08
10
11
12
13
14
9.342
10.34
11.34
12.34
13.34
10.47
11.53
12.58
13.64
14.69
11.78
12.90
14.01
15.12
16.22
13.44
14.63
15.81
16.98
18.15
15.99
17.28
18.55
19.81
21.06
18.31
19.68
21.03
22.36
23.68
20.48
21.92
23.34
24.74
26.12
23.21
24.72
26.22
27.69
29.14
25.19
26.76
28.30
29.82
31.32
29.59
31.26
32.91
34.53
36.12
31.42
33.14
34.82
36.48
38.11
15
16
17
18
19
3.108
3.536
3.980
4.439
4.912
3.483
3.942
4.416
4.905
5.407
4.601
5.142
5.697
6.265
6.844
5.229
5.812
6.408
7.015
7.633
6.262
6.908
7.564
8.231
8.907
7.261
7.962
8.672
9.390
10.12
8.547
9.312
10.09
10.86
11.65
10.31
11.15
12.00
12.86
13.72
11.72
12.62
13.53
14.44
15.35
13.03
13.98
14.94
15.89
16.85
15
16
17
18
19
14.34
15.34
16.34
17.34
18.34
15.73
16.78
17.82
18.87
19.91
17.32
18.42
19.51
20.60
21.69
19.31
20.47
21.61
22.76
23.90
22.31
23.54
24.77
25.99
27.20
25.00
26.30
27.59
28.87
30.14
27.49
28.85
30.19
31.53
32.85
30.58
32.00
33.41
34.81
36.19
32.80
34.27
35.72
37.16
38.58
37.70
39.25
40.79
42.31
43.82
39.72
41.31
42.88
44.43
45.97
20
21
22
23
24
5.398
5.896
6.404
6.924
7.453
5.921
6.447
6.983
7.529
8.085
7.434
8.034
8.643
9.260
9.886
8.260
8.897
9.542
10.20
10.86
9.591
10.28
10.98
11.69
12.40
10.85
11.59
12.34
13.09
13.85
12.44
13.24
14.04
14.85
15.66
14.58
15.44
16.31
17.19
18.06
16.27
17.18
18.10
19.02
19.94
17.81
18.77
19.73
20.69
21.65
20
21
22
23
24
19.34
20.34
21.34
22.34
23.34
20.95
21.99
23.03
24.07
25.11
22.77
23.86
24.94
26.02
27.10
25.04
26.17
27.30
28.43
29.55
28.41
29.62
30.81
32.01
33.20
31.41
32.67
33.92
35.17
36.42
34.17
35.48
36.78
38.08
39.36
37.57
38.93
40.29
41.64
42.98
40.00
41.40
42.80
44.18
45.56
45.31
46.80
48.27
49.73
51.18
47.50
49.01
50.51
52.00
53.48
25
26
27
28
29
7.991
8.538
9.093
9.656
10.23
8.649
9.222
9.803
10.39
10.99
10.52
11.16
11.81
12.46
13.12
11.52
12.20
12.88
13.56
14.26
13.12
13.84
14.57
15.31
16.05
14.61
15.38
16.15
16.93
17.71
16.47
17.29
18.11
18.94
19.77
18.94
19.82
20.70
21.59
22.48
20.87
21.79
22.72
23.65
24.58
22.62
23.58
24.54
25.51
26.48
25
26
27
28
29
24.34
25.34
26.34
27.34
28.34
26.14
27.18
28.21
29.25
30.28
28.17
29.25
30.32
31.39
32.46
30.68
31.79
32.91
34.03
35.14
34.38
35.56
36.74
37.92
39.09
37.65
38.89
40.11
41.34
42.56
40.65
41.92
43.19
44.46
45.72
44.31
45.64
46.96
48.28
49.59
46.93
48.29
49.64
50.99
52.34
52.62
54.05
55.48
56.89
58.30
54.95
56.41
57.86
59.30
60.73
30
32
34
36
38
10.80
11.98
13.18
14.40
15.64
11.59
12.81
14.06
15.32
16.61
13.79
15.13
16.50
17.89
19.29
14.95
16.36
17.79
19.23
20.69
16.79
18.29
19.81
21.34
22.88
18.49
20.07
21.66
23.27
24.88
20.60
22.27
23.95
25.64
27.34
23.36
25.15
26.94
28.73
30.54
25.51
27.37
29.24
31.12
32.99
27.44
29.38
31.31
33.25
35.19
30
32
34
36
38
29.34
31.34
33.34
35.34
37.34
31.32
33.38
35.44
37.50
39.56
33.53
35.66
37.80
39.92
42.05
36.25
38.47
40.68
42.88
45.08
40.26
42.58
44.90
47.21
49.51
43.77
46.19
48.60
51.00
53.38
46.98
49.48
51.97
54.44
56.90
50.89
53.49
56.06
58.62
61.16
53.67
56.33
58.96
61.58
64.18
59.70
62.49
65.25
67.99
70.70
62.16
65.00
67.80
70.59
73.35
40
50
60
70
80
16.91
23.46
30.34
37.47
44.79
17.92
24.67
31.74
39.04
46.52
20.71
27.99
35.53
43.28
51.17
22.16
29.71
37.48
45.44
53.54
24.43
32.36
40.48
48.76
57.15
26.51
34.76
43.19
51.74
60.39
29.05
37.69
46.46
55.33
64.28
32.34
41.45
50.64
59.90
69.21
34.87
44.31
53.81
63.35
72.92
37.13
46.86
56.62
66.40
76.19
40
50
60
70
80
39.34
49.33
59.33
69.33
79.33
41.62
51.89
62.13
72.36
82.57
44.16
54.72
65.23
75.69
86.12
47.27
58.16
68.97
79.71
90.41
51.81
63.17
74.40
85.53
96.58
55.76
67.50
79.08
90.53
101.9
59.34
71.42
83.30
95.02
106.6
63.69
76.15
88.38
100.4
112.3
66.77
79.49
91.95
104.2
116.3
73.40
86.66
99.61
112.3
124.8
76.09
89.56
102.7
115.6
128.3
90
100
52.28
59.90
54.1552
61.9179
59.20
67.33
61.75
70.06
65.65
74.22
69.13
77.93
73.29
82.36
78.56
87.95
82.51
92.13
85.99 90
95.81 100
89.33
99.33
92.76
102.9
96.52
106.9
101.1
111.7
107.6 113.1 118.1 124.1 128.3

118.5 124.3 129.6 135.8 140.2
137.2
149.4
140.8
153.2
Formulae and Tables
Page 1109 of 1113
5
6
7
8
9
ACTL2131 & ACTL5101
c Katja Ignatieva
!
P=
1
2
3
4
5% Points for the F distribution
10
12
24
161.4
18.51
10.13
7.709
199.5
19.00
9.552
6.944
215.7
19.16
9.277
6.591
224.6
19.25
9.117
6.388
230.2
19.30
9.013
6.256
234.0
19.33
8.941
6.163
236.8
19.35
8.887
6.094
238.9
19.37
8.845
6.041
240.5
19.38
8.812
5.999
241.9
19.40
8.786
5.964
243.9
19.41
8.745
5.912
249.1
19.45
8.639
5.774
254.3
19.50
8.526
5.628
3.191
2.818
2.575
2.404
2.277
3.105
2.722
2.471
2.293
2.159
5
6
7
8
9
6.608
5.987
5.591
5.318
5.117
5.786
5.143
4.737
4.459
4.256
5.409
4.757
4.347
4.066
3.863
5.192
4.534
4.120
3.838
3.633
5.050
4.387
3.972
3.687
3.482
4.950
4.284
3.866
3.581
3.374
4.876
4.207
3.787
3.500
3.293
4.818
4.147
3.726
3.438
3.230
4.772
4.099
3.677
3.388
3.179
4.735
4.060
3.637
3.347
3.137
4.678
4.000
3.575
3.284
3.073
4.527
3.841
3.410
3.115
2.900
4.365
3.669
3.230
2.928
2.707
2.284
2.209
2.147
2.097
2.054
2.178
2.100
2.036
1.983
1.938
2.055
1.972
1.904
1.846
1.797
10
11
12
13
14
4.965
4.844
4.747
4.667
4.600
4.103
3.982
3.885
3.806
3.739
3.708
3.587
3.490
3.411
3.344
3.478
3.357
3.259
3.179
3.112
3.326
3.204
3.106
3.025
2.958
3.217
3.095
2.996
2.915
2.848
3.135
3.012
2.913
2.832
2.764
3.072
2.948
2.849
2.767
2.699
3.020
2.896
2.796
2.714
2.646
2.978
2.854
2.753
2.671
2.602
2.913
2.788
2.687
2.604
2.534
2.737
2.609
2.505
2.420
2.349
2.538
2.404
2.296
2.206
2.131
2.059
2.028
2.001
1.977
1.956
2.017
1.985
1.958
1.933
1.912
1.899
1.866
1.836
1.810
1.787
1.755
1.718
1.686
1.657
1.631
15
16
17
18
19
4.543
4.494
4.451
4.414
4.381
3.682
3.634
3.592
3.555
3.522
3.287
3.239
3.197
3.160
3.127
3.056
3.007
2.965
2.928
2.895
2.901
2.852
2.810
2.773
2.740
2.790
2.741
2.699
2.661
2.628
2.707
2.657
2.614
2.577
2.544
2.641
2.591
2.548
2.510
2.477
2.588
2.538
2.494
2.456
2.423
2.544
2.494
2.450
2.412
2.378
2.475
2.425
2.381
2.342
2.308
2.288
2.235
2.190
2.150
2.114
2.066
2.010
1.960
1.917
1.878
1.965
1.948
1.933
1.919
1.906
1.937
1.920
1.904
1.890
1.877
1.892
1.875
1.859
1.845
1.832
1.767
1.748
1.731
1.716
1.702
1.607
1.586
1.567
1.549
1.533
20
21
22
23
24
4.351
4.325
4.301
4.279
4.260
3.493
3.467
3.443
3.422
3.403
3.098
3.072
3.049
3.028
3.009
2.866
2.840
2.817
2.796
2.776
2.711
2.685
2.661
2.640
2.621
2.599
2.573
2.549
2.528
2.508
2.514
2.488
2.464
2.442
2.423
2.447
2.420
2.397
2.375
2.355
2.393
2.366
2.342
2.320
2.300
2.348
2.321
2.297
2.275
2.255
2.278
2.250
2.226
2.204
2.183
2.082
2.054
2.028
2.005
1.984
1.843
1.812
1.783
1.757
1.733
1.929
1.919
1.909
1.900
1.892
1.895
1.884
1.874
1.865
1.857
1.866
1.855
1.845
1.836
1.827
1.820
1.809
1.799
1.790
1.781
1.689
1.677
1.666
1.656
1.647
1.518
1.504
1.491
1.478
1.467
25
26
27
28
29
4.242
4.225
4.210
4.196
4.183
3.385
3.369
3.354
3.340
3.328
2.991
2.975
2.960
2.947
2.934
2.759
2.743
2.728
2.714
2.701
2.603
2.587
2.572
2.558
2.545
2.490
2.474
2.459
2.445
2.432
2.405
2.388
2.373
2.359
2.346
2.337
2.321
2.305
2.291
2.278
2.282
2.265
2.250
2.236
2.223
2.236
2.220
2.204
2.190
2.177
2.165
2.148
2.132
2.118
2.104
1.964
1.946
1.930
1.915
1.901
1.711
1.691
1.672
1.654
1.638
1.927
1.913
1.901
1.891
1.881
1.884
1.870
1.858
1.847
1.838
1.849
1.835
1.822
1.811
1.802
1.819
1.805
1.793
1.781
1.772
1.773
1.758
1.745
1.734
1.724
1.638
1.622
1.608
1.595
1.584
1.456
1.437
1.419
1.404
1.390
30
32
34
36
38
4.171
4.149
4.130
4.113
4.098
3.316
3.295
3.276
3.259
3.245
2.922
2.901
2.883
2.866
2.852
2.690
2.668
2.650
2.634
2.619
2.534
2.512
2.494
2.477
2.463
2.421
2.399
2.380
2.364
2.349
2.334
2.313
2.294
2.277
2.262
2.266
2.244
2.225
2.209
2.194
2.211
2.189
2.170
2.153
2.138
2.165
2.142
2.123
2.106
2.091
2.092
2.070
2.050
2.033
2.017
1.887
1.864
1.843
1.824
1.808
1.622
1.594
1.569
1.547
1.527
1.873
1.819
1.767
1.717
1.829
1.775
1.722
1.670
1.793
1.738
1.684
1.632
1.763
1.707
1.652
1.599
1.715
1.657
1.601
1.546
1.574
1.511
1.447
1.383
1.377
1.291
1.193
1.000
40
60
120
4.085
4.001
3.920
3.841
3.232
3.150
3.072
2.996
2.839
2.758
2.680
2.605
2.606
2.525
2.447
2.372
2.449
2.368
2.290
2.214
2.336
2.254
2.175
2.099
2.249
2.167
2.087
2.010
2.180
2.097
2.016
1.938
2.124
2.040
1.959
1.880
2.077
1.993
1.910
1.831
2.003
1.917
1.834
1.752
1.793
1.700
1.608
1.517
1.509
1.389
1.254
1.000
10
12
24
39.86
8.526
5.538
4.545
49.50
9.000
5.462
4.325
53.59
9.162
5.391
4.191
55.83
9.243
5.343
4.107
57.24
9.293
5.309
4.051
58.20
9.326
5.285
4.010
58.91
9.349
5.266
3.979
59.44
9.367
5.252
3.955
59.86
9.381
5.240
3.936
60.19
9.392
5.230
3.920
60.71
9.408
5.216
3.896
62.00
9.450
5.176
3.831
5
6
7
8
9
4.060
3.776
3.589
3.458
3.360
3.780
3.463
3.257
3.113
3.006
3.619
3.289
3.074
2.924
2.813
3.520
3.181
2.961
2.806
2.693
3.453
3.108
2.883
2.726
2.611
3.405
3.055
2.827
2.668
2.551
3.368
3.014
2.785
2.624
2.505
3.339
2.983
2.752
2.589
2.469
3.316
2.958
2.725
2.561
2.440
3.297
2.937
2.703
2.538
2.416
3.268
2.905
2.668
2.502
2.379
10
11
12
13
14
3.285
3.225
3.177
3.136
3.102
2.924
2.860
2.807
2.763
2.726
2.728
2.660
2.606
2.560
2.522
2.605
2.536
2.480
2.434
2.395
2.522
2.451
2.394
2.347
2.307
2.461
2.389
2.331
2.283
2.243
2.414
2.342
2.283
2.234
2.193
2.377
2.304
2.245
2.195
2.154
2.347
2.274
2.214
2.164
2.122
2.323
2.248
2.188
2.138
2.095
15
16
17
18
19
3.073
3.048
3.026
3.007
2.990
2.695
2.668
2.645
2.624
2.606
2.490
2.462
2.437
2.416
2.397
2.361
2.333
2.308
2.286
2.266
2.273
2.244
2.218
2.196
2.176
2.208
2.178
2.152
2.130
2.109
2.158
2.128
2.102
2.079
2.058
2.119
2.088
2.061
2.038
2.017
2.086
2.055
2.028
2.005
1.984
20
21
22
23
24
2.975
2.961
2.949
2.937
2.927
2.589
2.575
2.561
2.549
2.538
2.380
2.365
2.351
2.339
2.327
2.249
2.233
2.219
2.207
2.195
2.158
2.142
2.128
2.115
2.103
2.091
2.075
2.060
2.047
2.035
2.040
2.023
2.008
1.995
1.983
1.999
1.982
1.967
1.953
1.941
25
26
27
28
29
2.918
2.909
2.901
2.894
2.887
2.528
2.519
2.511
2.503
2.495
2.317
2.307
2.299
2.291
2.283
2.184
2.174
2.165
2.157
2.149
2.092
2.082
2.073
2.064
2.057
2.024
2.014
2.005
1.996
1.988
1.971
1.961
1.952
1.943
1.935
30
32
34
36
38
2.881
2.869
2.859
2.850
2.842
2.489
2.477
2.466
2.456
2.448
2.276
2.263
2.252
2.243
2.234
2.142
2.129
2.118
2.108
2.099
2.049
2.036
2.024
2.014
2.005
1.980
1.967
1.955
1.945
1.935
40
60
120
2.835
2.791
2.748
2.706
2.440
2.393
2.347
2.303
2.226
2.177
2.130
2.084
2.091
2.041
1.992
1.945
1.997
1.946
1.896
1.847
1.927
1.875
1.824
1.774
Formulae and Tables
Page 1110 of 1113
ACTL2131 & ACTL5101
c Katja Ignatieva
!
63.33
9.491
5.134
3.761
1 =
2
1
2
3
4
1 =
2
1
2
3
4
2.5% Points for the F distribution
10
12
24
4052
98.50
34.12
21.20
4999
99.00
30.82
18.00
5403
99.17
29.46
16.69
5625
99.25
28.71
15.98
5764
99.30
28.24
15.52
5859
99.33
27.91
15.21
5928
99.36
27.67
14.98
5981
99.37
27.49
14.80
6022
99.39
27.35
14.66
6056
99.40
27.23
14.55
6106
99.42
27.05
14.37
6235
99.46
26.60
13.93
6366
99.50
26.13
13.46
6.278
5.117
4.415
3.947
3.614
6.015
4.849
4.142
3.670
3.333
5
6
7
8
9
16.26
13.75
12.25
11.26
10.56
13.27
10.92
9.547
8.649
8.022
12.06
9.780
8.451
7.591
6.992
11.39
9.148
7.847
7.006
6.422
10.97
8.746
7.460
6.632
6.057
10.67
8.466
7.191
6.371
5.802
10.46
8.260
6.993
6.178
5.613
10.29
8.102
6.840
6.029
5.467
10.16
7.976
6.719
5.911
5.351
10.05
7.874
6.620
5.814
5.257
9.888
7.718
6.469
5.667
5.111
9.466
7.313
6.074
5.279
4.729
9.020
6.880
5.650
4.859
4.311
3.621
3.430
3.277
3.153
3.050
3.365
3.173
3.019
2.893
2.789
3.080
2.883
2.725
2.595
2.487
10
11
12
13
14
10.04
9.646
9.330
9.074
8.862
7.559
7.206
6.927
6.701
6.515
6.552
6.217
5.953
5.739
5.564
5.994
5.668
5.412
5.205
5.035
5.636
5.316
5.064
4.862
4.695
5.386
5.069
4.821
4.620
4.456
5.200
4.886
4.640
4.441
4.278
5.057
4.744
4.499
4.302
4.140
4.942
4.632
4.388
4.191
4.030
4.849
4.539
4.296
4.100
3.939
4.706
4.397
4.155
3.960
3.800
4.327
4.021
3.780
3.587
3.427
3.909
3.602
3.361
3.165
3.004
3.060
2.986
2.922
2.866
2.817
2.963
2.889
2.825
2.769
2.720
2.701
2.625
2.560
2.503
2.452
2.395
2.316
2.247
2.187
2.133
15
16
17
18
19
8.683
8.531
8.400
8.285
8.185
6.359
6.226
6.112
6.013
5.926
5.417
5.292
5.185
5.092
5.010
4.893
4.773
4.669
4.579
4.500
4.556
4.437
4.336
4.248
4.171
4.318
4.202
4.102
4.015
3.939
4.142
4.026
3.927
3.841
3.765
4.004
3.890
3.791
3.705
3.631
3.895
3.780
3.682
3.597
3.523
3.805
3.691
3.593
3.508
3.434
3.666
3.553
3.455
3.371
3.297
3.294
3.181
3.084
2.999
2.925
2.868
2.753
2.653
2.566
2.489
2.837
2.798
2.763
2.731
2.703
2.774
2.735
2.700
2.668
2.640
2.676
2.637
2.602
2.570
2.541
2.408
2.368
2.331
2.299
2.269
2.085
2.042
2.003
1.968
1.935
20
21
22
23
24
8.096
8.017
7.945
7.881
7.823
5.849
5.780
5.719
5.664
5.614
4.938
4.874
4.817
4.765
4.718
4.431
4.369
4.313
4.264
4.218
4.103
4.042
3.988
3.939
3.895
3.871
3.812
3.758
3.710
3.667
3.699
3.640
3.587
3.539
3.496
3.564
3.506
3.453
3.406
3.363
3.457
3.398
3.346
3.299
3.256
3.368
3.310
3.258
3.211
3.168
3.231
3.173
3.121
3.074
3.032
2.859
2.801
2.749
2.702
2.659
2.421
2.360
2.305
2.256
2.211
2.753
2.729
2.707
2.687
2.669
2.677
2.653
2.631
2.611
2.592
2.613
2.590
2.568
2.547
2.529
2.515
2.491
2.469
2.448
2.430
2.242
2.217
2.195
2.174
2.154
1.906
1.878
1.853
1.829
1.807
25
26
27
28
29
7.770
7.721
7.677
7.636
7.598
5.568
5.526
5.488
5.453
5.420
4.675
4.637
4.601
4.568
4.538
4.177
4.140
4.106
4.074
4.045
3.855
3.818
3.785
3.754
3.725
3.627
3.591
3.558
3.528
3.499
3.457
3.421
3.388
3.358
3.330
3.324
3.288
3.256
3.226
3.198
3.217
3.182
3.149
3.120
3.092
3.129
3.094
3.062
3.032
3.005
2.993
2.958
2.926
2.896
2.868
2.620
2.585
2.552
2.522
2.495
2.169
2.131
2.097
2.064
2.034
2.746
2.715
2.688
2.664
2.643
2.651
2.620
2.593
2.569
2.548
2.575
2.543
2.516
2.492
2.471
2.511
2.480
2.453
2.429
2.407
2.412
2.381
2.353
2.329
2.307
2.136
2.103
2.075
2.049
2.027
1.787
1.750
1.717
1.687
1.661
30
32
34
36
38
7.562
7.499
7.444
7.396
7.353
5.390
5.336
5.289
5.248
5.211
4.510
4.459
4.416
4.377
4.343
4.018
3.969
3.927
3.890
3.858
3.699
3.652
3.611
3.574
3.542
3.473
3.427
3.386
3.351
3.319
3.304
3.258
3.218
3.183
3.152
3.173
3.127
3.087
3.052
3.021
3.067
3.021
2.981
2.946
2.915
2.979
2.934
2.894
2.859
2.828
2.843
2.798
2.758
2.723
2.692
2.469
2.423
2.383
2.347
2.316
2.006
1.956
1.911
1.872
1.837
2.624
2.507
2.395
2.288
2.529
2.412
2.299
2.192
2.452
2.334
2.222
2.114
2.388
2.270
2.157
2.048
2.288
2.169
2.055
1.945
2.007
1.882
1.760
1.640
1.637
1.482
1.310
1.000
40
60
120
7.314
7.077
6.851
6.635
5.179
4.977
4.787
4.605
4.313
4.126
3.949
3.782
3.828
3.649
3.480
3.319
3.514
3.339
3.174
3.017
3.291
3.119
2.956
2.802
3.124
2.953
2.792
2.639
2.993
2.823
2.663
2.511
2.888
2.718
2.559
2.407
2.801
2.632
2.472
2.321
2.665
2.496
2.336
2.185
2.288
2.115
1.950
1.791
1.805
1.601
1.381
1.000
10
12
24
647.8
38.51
17.44
12.22
799.5
39.00
16.04
10.65
864.2
39.17
15.44
9.979
899.6
39.25
15.10
9.605
921.8
39.30
14.88
9.364
937.1
39.33
14.73
9.197
948.2
39.36
14.62
9.074
956.7
39.37
14.54
8.980
963.3
39.39
14.47
8.905
968.6
39.40
14.42
8.844
976.7
39.41
14.34
8.751
997.2
39.46
14.12
8.511
5
6
7
8
9
10.01
8.813
8.073
7.571
7.209
8.434
7.260
6.542
6.059
5.715
7.764
6.599
5.890
5.416
5.078
7.388
6.227
5.523
5.053
4.718
7.146
5.988
5.285
4.817
4.484
6.978
5.820
5.119
4.652
4.320
6.853
5.695
4.995
4.529
4.197
6.757
5.600
4.899
4.433
4.102
6.681
5.523
4.823
4.357
4.026
6.619
5.461
4.761
4.295
3.964
6.525
5.366
4.666
4.200
3.868
10
11
12
13
14
6.937
6.724
6.554
6.414
6.298
5.456
5.256
5.096
4.965
4.857
4.826
4.630
4.474
4.347
4.242
4.468
4.275
4.121
3.996
3.892
4.236
4.044
3.891
3.767
3.663
4.072
3.881
3.728
3.604
3.501
3.950
3.759
3.607
3.483
3.380
3.855
3.664
3.512
3.388
3.285
3.779
3.588
3.436
3.312
3.209
3.717
3.526
3.374
3.250
3.147
15
16
17
18
19
6.200
6.115
6.042
5.978
5.922
4.765
4.687
4.619
4.560
4.508
4.153
4.077
4.011
3.954
3.903
3.804
3.729
3.665
3.608
3.559
3.576
3.502
3.438
3.382
3.333
3.415
3.341
3.277
3.221
3.172
3.293
3.219
3.156
3.100
3.051
3.199
3.125
3.061
3.005
2.956
3.123
3.049
2.985
2.929
2.880
20
21
22
23
24
5.871
5.827
5.786
5.750
5.717
4.461
4.420
4.383
4.349
4.319
3.859
3.819
3.783
3.750
3.721
3.515
3.475
3.440
3.408
3.379
3.289
3.250
3.215
3.183
3.155
3.128
3.090
3.055
3.023
2.995
3.007
2.969
2.934
2.902
2.874
2.913
2.874
2.839
2.808
2.779
25
26
27
28
29
5.686
5.659
5.633
5.610
5.588
4.291
4.265
4.242
4.221
4.201
3.694
3.670
3.647
3.626
3.607
3.353
3.329
3.307
3.286
3.267
3.129
3.105
3.083
3.063
3.044
2.969
2.945
2.923
2.903
2.884
2.848
2.824
2.802
2.782
2.763
30
32
34
36
38
5.568
5.531
5.499
5.471
5.446
4.182
4.149
4.120
4.094
4.071
3.589
3.557
3.529
3.505
3.483
3.250
3.218
3.191
3.167
3.145
3.026
2.995
2.968
2.944
2.923
2.867
2.836
2.808
2.785
2.763
40
60
120
5.424
5.286
5.152
5.024
4.051
3.925
3.805
3.689
3.463
3.343
3.227
3.116
3.126
3.008
2.894
2.786
2.904
2.786
2.674
2.566
2.744
2.627
2.515
2.408
Formulae and Tables
Page 1111 of 1113
ACTL2131 & ACTL5101
c Katja Ignatieva
!
1018
39.50
13.90
8.257
1 =
2
1
2
3
4
1 =
2
1
2
3
4
Stochastic processes
Markov jump processes
Kolmogorov differential equations 8
pik (s, t) kj (t)

Forward equation:
t pij (s, t) =
kS
8
Backward equation:
ik (s) pkj (s, t)
t pij (s, t) =
kS
where ij (t) is the transition

rate from state i to state j (j .= i) at
8
time t, and ii = ij .
j'=i
Expected time8to reach a subsequent

8 state k
ij
mi = 1i +
ij
i mj , where i =
j'=i,j'=k
7.2
j'=i
Brownian motion and related processes

Martingales for standard Brownian motion
If {Bt , t 0} is a standard Brownian motion, then the following
processes are martingales:
Bt , Bt2 t and exp(Bt 12 2 t)
Distribution
of the maximum
value
+
,
<
=
<
=
yt
2y
P max (Bs + s) > y = y+t

+
e
,y>0
t
t
0st
7.3
Hitting times
If y = min{s : Bs + s = y} where > 0 and y < 0, then
s0
2
E[ey ] = ey(+ +2 , > 0
Ornstein-Uhlenbeck process
dXt = Xt dt + dBt , > 0
Monte Carlo methods
Box-Muller formulae
If U1 and U2 are independent random variables form the U (0, 1)
distribution
then
Z1 = 2 log U1 cos(2U2 ) and Z2 = 2 log U1 sin(2U2 )

are independent standard normal variables.
Polar method
If V1 and V2 are independent random variables from the U (1, 1)
distributionMand S = V12 + V22 then,
M conditional on 0 < S 1,
ACTL2131 & ACTL5101
c Katja Ignatieva
!
7
7.1
S
S
and Z2 = V2 2 log
Z1 = V1 2 log
S
S
are independent standard normal variables.
Psuedorandom values from the U (0, 1) distribution and the N (0, 1)
distribution are included in the statistical tables section.
Formulae and Tables
Page 1112 of 1113
Time series
Time series - Time domain
Sample autocovariance and autocorrelation function
n
8
(xt
6) (xtk
6), where
6=
Autocovariance:
6k = n1
t=k+1
$
k
0
$
Autocorrelation: 6k =
Autocorrelation function for ARMA(1,1)
For the process Xt = Xt1 + et + et1 :
k1
k = (1+)(+)
, k = 1, 2, 3, . . .
1+ 2 +2
Partial autocorrelation function
2 21
1 = 1 2 = 1
2
k =
det(Pk$ )
,
det(P
k )
k = 2, 3, . . .,
1
1
2
. . . k1
1
1
1
. . . k2
2
1
1
. . . k3
where Pk =
..
..
..
..
.
.
.
.
.
.
.
k1 k2 k3 . . .
1
and Pk) equals Pk , but with the last column replaced with
$
[1 , 2 , 3 , . . . , k ] .
8.3
k=1
Page 1113 of 1113
where rk (k = 1, 2, . . . , m0 is the estimated value of the k th

autocorrelation coefficient of the residuals and n is the number of
data values used inn the ARMA(p, q) series
Turning point test
In a sequence of n independent random variables the number of
turning points T is such that:
E[T ] = 32 (n 2) and var(T ) = 16n29
90
n
8
t=1
xt
8.2
k=
Inversion formula
H
k = eik f ()d
Spectral density function for ARMA(p, q)

The spectral density function of the process (B)(Xt ) = (B)et ,
where var(et ) = 2 , is
2 (ei )(ei )
f () = 2 (e
i )(ei )
Linear filters
8
For the linear filter Yt =
ak Xtk :
2
k=
fY () = |A()| fX ()
8
where A() =
eik ak is the transfer function for the filter.
k=
Formulae and Tables
Time series - Box-Jenkins methodology

Ljung and Box Portmanteau test of the residuals for an
ARMA(p, q) model
m
2
8
rk
2
n(n + 2)
nk m(p+q)
1
n
Partial autocorrelation function for MA(1)

For the process Xt = + et + et1 :
(1 2 ) k
k = (1)k+1 1
2(k+1) , k = 1, 2, 3, . . .
Time series - Frequency domain
Spectral density function
8
1
f () = 2
eik k , < <
ACTL2131 & ACTL5101
c Katja Ignatieva
!
8
8.1

Actuarial Questions

Uploaded by

Copyright:

Available Formats

Actuarial Questions

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Actuarial Questions

Uploaded by

Copyright:

Available Formats

What is the purpose of tutorials according to the document?

What is the purpose of tutorials according to the document?

What are the three groups of exercises covered in tutorials?

What are the three groups of exercises covered in tutorials?

Title page

Australian School of Business

Probability and Statistics

Australian School of Business

School of Risk and Actuarial Studies, ASB, UNSW

ACTL2131 & ACTL5101

Probability and Statistics

Figure 1: Feedback loop

School of Risk and Actuarial Studies, ASB, UNSW

ACTL2131 & ACTL5101

Probability and Statistics

10, 11, 12, 13

14, 15, 16, 17

Revision for exam

Table 1: Schedule of tutorial exercises.

School of Risk and Actuarial Studies, ASB, UNSW

Page iii of iii

Australian School of Business

= {the sum is between (and includes) 7 and 10}

Show that E1 , E2 and E3 are independent.

6. In an undergraduate statistics class, three students A, B, and C submitted exactly (word-for-word)

School of Risk and Actuarial Studies, ASB, UNSW

ACTL2131 & ACTL5101

Probability and Statistics

Tutorial Exercises, Week 1

x > 1, and zero otherwise.

(b) Sketch the graphs of pX (x) and FX (x).

, for < x < .

Here, X is called a normally distributed random variable.

School of Risk and Actuarial Studies, ASB, UNSW

ACTL2131 & ACTL5101

Probability and Statistics

Tutorial Exercises, Week 1

(d) How do we call the function S (t)?

Verify that fX () is a pdf.

where FX () denotes the cdf of X.

(b) Show that for a non-negative random variable:

Use this result to show that:

School of Risk and Actuarial Studies, ASB, UNSW

ACTL2131 & ACTL5101

Probability and Statistics

Tutorial Exercises, Week 1

17. A random variable X has a probability density function of the form:

(a) Determine the value of a in terms of b and show that b 1.

(b) For the case b = 1, determine the mean and variance of X.

-End of week 1 Tutorial Exercises-

School of Risk and Actuarial Studies, ASB, UNSW

Australian School of Business

School of Risk and Actuarial Studies, ASB, UNSW

ACTL2131 & ACTL5101

Probability and Statistics

Tutorial Exercises, Week 2

(a) Prove that limn MX (t) = MY (t).

Second, a similar relation can be approximated for the binomial distribution:

Hint: first show that