6/4/2018 Coursera | Online Courses From Top Universities.

Join for Free | Coursera

Week 1 Lecture Notes

ML:Introduction
What is Machine Learning?
Two de nitions of Machine Learning are o ered. Arthur Samuel described it as: "the eld of study that gives
computers the ability to learn without being explicitly programmed." This is an older, informal de nition.

Tom Mitchell provides a more modern de nition: "A computer program is said to learn from experience E
with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as
measured by P, improves with experience E."

Example: playing checkers.

E = the experience of playing many games of checkers

T = the task of playing checkers.

P = the probability that the program will win the next game.

In general, any machine learning problem can be assigned to one of two broad classi cations:

supervised learning, OR

unsupervised learning.

Supervised Learning

In supervised learning, we are given a data set and already know what our correct output should look like,
having the idea that there is a relationship between the input and the output.

Supervised learning problems are categorized into "regression" and "classi cation" problems. In a
regression problem, we are trying to predict results within a continuous output, meaning that we are trying
to map input variables to some continuous function. In a classi cation problem, we are instead trying to
predict results in a discrete output. In other words, we are trying to map input variables into discrete
categories. Here is a description on Math is Fun on Continuous and Discrete Data.

Example 1:

https://www.coursera.org/learn/machine-learning/resources/JXWWS 1/8
6/4/2018 Coursera | Online Courses From Top Universities. Join for Free | Coursera

Given data about the size of houses on the real estate market, try to predict their price. Price as a function
of size is a continuous output, so this is a regression problem.

We could turn this example into a classi cation problem by instead making our output about whether the
house "sells for more or less than the asking price." Here we are classifying the houses based on price into
two discrete categories.

Example 2:

(a) Regression - Given a picture of Male/Female, We have to predict his/her age on the basis of given
picture.

(b) Classi cation - Given a picture of Male/Female, We have to predict Whether He/She is of High school,
College, Graduate age. Another Example for Classi cation - Banks have to decide whether or not to give a
loan to someone on the basis of his credit history.

Unsupervised Learning

Unsupervised learning, on the other hand, allows us to approach problems with little or no idea what our
results should look like. We can derive structure from data where we don't necessarily know the e ect of
the variables.

We can derive this structure by clustering the data based on relationships among the variables in the data.

With unsupervised learning there is no feedback based on the prediction results, i.e., there is no teacher to
correct you.

Example:

Clustering: Take a collection of 1000 essays written on the US Economy, and nd a way to automatically
group these essays into a small number that are somehow similar or related by di erent variables, such as
word frequency, sentence length, page count, and so on.

Non-clustering: The "Cocktail Party Algorithm", which can nd structure in messy data (such as the
identi cation of individual voices and music from a mesh of sounds at a cocktail party
(https://en.wikipedia.org/wiki/Cocktail_party_e ect) ). Here is an answer on Quora to enhance your
understanding. : https://www.quora.com/What-is-the-di erence-between-supervised-and-unsupervised-
learning-algorithms ?

ML:Linear Regression with One Variable

Model Representation
Recall that in regression problems, we are taking input variables and trying to t the output onto a
continuous expected result function.

https://www.coursera.org/learn/machine-learning/resources/JXWWS 2/8
6/4/2018 Coursera | Online Courses From Top Universities. Join for Free | Coursera

Linear regression with one variable is also known as "univariate linear regression."

Univariate linear regression is used when you want to predict a single output value y from a single input
value x. We're doing supervised learning here, so that means we already have an idea about what the
input/output cause and e ect should be.

The Hypothesis Function

Our hypothesis function has the general form:

^
y = h θ (x) = θ 0 + θ 1 x

Note that this is like the equation of a straight line. We give to hθ (x) values for θ 0 and θ 1 to get our
estimated output ^ y . In other words, we are trying to create a function called h θ that is trying to map our

input data (the x's) to our output data (the y's).

Example:

Suppose we have the following set of training data:

input x output y

0 4

1 7

2 7

3 8

Now we can make a random guess about our hθ function: θ 0 = 2 and θ 1 = 2 . The hypothesis function
becomes hθ (x) = 2 + 2x .

So for input of 1 to our hypothesis, y will be 4. This is o by 3. Note that we will be trying out various values
of θ 0 and θ 1 to try to nd values which provide the best possible " t" or the most representative "straight
line" through the data points mapped on the x-y plane.

Cost Function
We can measure the accuracy of our hypothesis function by using a cost function. This takes an average
(actually a fancier version of an average) of all the results of the hypothesis with inputs from x's compared
to the actual output y's.

m m
2
1 1 2
^
J(θ 0 , θ 1 ) = y − yi = (h θ (xi ) − y i )
( i )
2m ∑ 2m ∑
i=1 i=1

https://www.coursera.org/learn/machine-learning/resources/JXWWS 3/8
6/4/2018 Coursera | Online Courses From Top Universities. Join for Free | Coursera

To break it apart, it is where x̄ is the mean of the squares of hθ (xi ) − yi , or the di erence between the
1
x̄
2

predicted value and the actual value.

This function is otherwise called the "Squared error function", or "Mean squared error". The mean is halved
) as a convenience for the computation of the gradient descent, as the derivative term of the square
1
( 2m

function will cancel out the 1

2
term.

Now we are able to concretely measure the accuracy of our predictor function against the correct results
we have so that we can predict new results we don't have.

If we try to think of it in visual terms, our training data set is scattered on the x-y plane. We are trying to
make straight line (de ned by hθ (x)) which passes through this scattered set of data. Our objective is to get
the best possible line. The best possible line will be such so that the average squared vertical distances of
the scattered points from the line will be the least. In the best case, the line should pass through all the
points of our training data set. In such a case the value of J(θ 0 , θ 1 ) will be 0.

ML:Gradient Descent
So we have our hypothesis function and we have a way of measuring how well it ts into the data. Now we
need to estimate the parameters in hypothesis function. That's where gradient descent comes in.

Imagine that we graph our hypothesis function based on its elds θ 0 and θ 1 (actually we are graphing the
cost function as a function of the parameter estimates). This can be kind of confusing; we are moving up to
a higher level of abstraction. We are not graphing x and y itself, but the parameter range of our hypothesis
function and the cost resulting from selecting particular set of parameters.

We put θ 0 on the x axis and θ 1 on the y axis, with the cost function on the vertical z axis. The points on our
graph will be the result of the cost function using our hypothesis with those speci c theta parameters.

We will know that we have succeeded when our cost function is at the very bottom of the pits in our graph,
i.e. when its value is the minimum.

The way we do this is by taking the derivative (the tangential line to a function) of our cost function. The
slope of the tangent is the derivative at that point and it will give us a direction to move towards. We make
steps down the cost function in the direction with the steepest descent, and the size of each step is
determined by the parameter α, which is called the learning rate.

The gradient descent algorithm is:

repeat until convergence:

∂
θ j := θ j − α J(θ 0 , θ 1 )
∂θj

where

j=0,1 represents the feature index number.

https://www.coursera.org/learn/machine-learning/resources/JXWWS 4/8
6/4/2018 Coursera | Online Courses From Top Universities. Join for Free | Coursera

Intuitively, this could be thought of as:

repeat until convergence:

θ j := θ j − α[Slope of tangent aka derivative in j dimension] [Slope of tangent aka derivative in j

dimension]

Gradient Descent for Linear Regression

When speci cally applied to the case of linear regression, a new form of the gradient descent equation can
be derived. We can substitute our actual cost function and our actual hypothesis function and modify the
equation to (the derivation of the formulas are out of the scope of this course, but a really great one can be
found here):

repeat until convergence: {

m
1
θ 0 := θ 0 − α (h θ (x i ) − y i )
m ∑
i=1

m
1
θ 1 := θ 1 − α ((h θ (x i ) − y i )x i )
m ∑
i=1

where m is the size of the training set, θ 0 a constant that will be changing simultaneously with θ 1 and
xi , y i are values of the given training set (data).

Note that we have separated out the two cases for θ j into separate equations for θ 0 and θ 1 ; and that for θ 1
we are multiplying xi at the end due to the derivative.

The point of all this is that if we start with a guess for our hypothesis and then repeatedly apply these
gradient descent equations, our hypothesis will become more and more accurate.

Gradient Descent for Linear Regression: visual worked example

Some may nd the following video (https://www.youtube.com/watch?v=WnqQrPNYz5Q) useful as it

visualizes the improvement of the hypothesis as the error function reduces.

ML:Linear Algebra Review

Khan Academy has excellent Linear Algebra Tutorials (https://www.khanacademy.org/#linear-algebra)

Matrices and Vectors

Matrices are 2-dimensional arrays:
https://www.coursera.org/learn/machine-learning/resources/JXWWS 5/8
6/4/2018 Coursera | Online Courses From Top Universities. Join for Free | Coursera

⎡a b c⎤
⎢ ⎥
d e f
⎢ ⎥
g h i
⎢ ⎥
⎣ j k l ⎦

The above matrix has four rows and three columns, so it is a 4 x 3 matrix.

A vector is a matrix with one column and many rows:

⎡w⎤
⎢ ⎥
x
⎢ ⎥
⎢ y ⎥
⎣ z ⎦

So vectors are a subset of matrices. The above vector is a 4 x 1 matrix.

Notation and terms:

Aij refers to the element in the ith row and jth column of matrix A.

A vector with 'n' rows is referred to as an 'n'-dimensional vector

vi refers to the element in the ith row of the vector.

In general, all our vectors and matrices will be 1-indexed. Note that for some programming languages, the
arrays are 0-indexed.

Matrices are usually denoted by uppercase names while vectors are lowercase.

"Scalar" means that an object is a single value, not a vector or matrix.

ℝ refers to the set of scalar real numbers

ℝ refers to the set of n-dimensional vectors of real numbers

Addition and Scalar Multiplication

Addition and subtraction are element-wise, so you simply add or subtract each corresponding element:

a b w x a + w b + x
+ =
[c d] [ y z] [ c + y d + z]

To add or subtract two matrices, their dimensions must be the same.

https://www.coursera.org/learn/machine-learning/resources/JXWWS 6/8
6/4/2018 Coursera | Online Courses From Top Universities. Join for Free | Coursera

In scalar multiplication, we simply multiply every element by the scalar value:

a b a ∗ x b ∗ x
∗ x =
[c d] [c ∗ x d ∗ x]

Matrix-Vector Multiplication
We map the column of the vector onto each row of the matrix, multiplying each element and summing the
result.

⎡a ⎡ a ∗ x + b ∗ y⎤
b⎤
⎢ ⎥ x ⎢ ⎥
⎢c d⎥ ∗ = c ∗ x + d ∗ y
[y] ⎢ ⎥
⎣e f ⎦ ⎣ e ∗ x + f ∗ y⎦

The result is a vector. The vector must be the second term of the multiplication. The number of columns
of the matrix must equal the number of rows of the vector.

An m x n matrix multiplied by an n x 1 vector results in an m x 1 vector.

Matrix-Matrix Multiplication
We multiply two matrices by breaking it into several vector multiplications and concatenating the result

⎡a ⎡a ∗ w + b ∗ y a ∗ x + b ∗ z⎤
b⎤
⎢ ⎥ w x ⎢ ⎥
⎢c d⎥ ∗ = c ∗ w + d ∗ y c ∗ x + d ∗ z
[ y ⎢ ⎥
z]
⎣e f ⎦ ⎣ e ∗ w + f ∗ y e ∗ x + f ∗ z⎦

An m x n matrix multiplied by an n x o matrix results in an m x o matrix. In the above example, a 3 x 2

matrix times a 2 x 2 matrix resulted in a 3 x 2 matrix.

To multiply two matrices, the number of columns of the rst matrix must equal the number of rows of the
second matrix.

Matrix Multiplication Properties

Not commutative. A∗B≠B∗A

Associative. (A∗B)∗C=A∗(B∗C)

https://www.coursera.org/learn/machine-learning/resources/JXWWS 7/8
6/4/2018 Coursera | Online Courses From Top Universities. Join for Free | Coursera

The identity matrix, when multiplied by any matrix of the same dimensions, results in the original matrix.
It's just like multiplying numbers by 1. The identity matrix simply has 1's on the diagonal (upper left to lower
right diagonal) and 0's elsewhere.

⎡1 0 0⎤
⎢ ⎥
⎢0 1 0⎥

⎣0 0 1⎦

When multiplying the identity matrix after some matrix (A∗I), the square identity matrix should match the
other matrix's columns. When multiplying the identity matrix before some other matrix (I∗A), the square
identity matrix should match the other matrix's rows.

Inverse and Transpose

The inverse of a matrix A is denoted A−1. Multiplying by the inverse results in the identity matrix.

A non square matrix does not have an inverse matrix. We can compute inverses of matrices in octave with
the pinv(A) function [1] and in matlab with the inv(A) function. Matrices that don't have an inverse are
singular or degenerate.

The transposition of a matrix is like rotating the matrix 90° in clockwise direction and then reversing it. We
can compute transposition of matrices in matlab with the transpose(A) function or A':

⎡a b⎤
⎢ ⎥
A =
⎢c d⎥

⎣e f ⎦

T a c e
A =
[b d f ]

In other words:

T
Aij = Aji

https://www.coursera.org/learn/machine-learning/resources/JXWWS 8/8