Algorithm

Chapter 6
Dynamic Programming
CS 350: Winter 2018
1
Algorithmic Paradigms
Greedy. Build up a solution incrementally, myopically optimizing some local
criterion.
Divide-and-conquer. Break up a problem into sub-problems, solve each sub-

problem independently, and combine solution to sub-problems to form solution to
original problem.
Dynamic programming. Break up a problem into a series of overlapping sub-

problems, and build up solutions to larger and larger sub-problems; typically each
subproblem is solved just once, and the solution is stored (i.e. cached).
The next time the same subproblem occurs, instead of recomputing its solution,
one simply looks up the previously computed solution, thereby saving computation
time at the expense of a (hopefully) modest expenditure in storage space. Each
of the subproblem solutions is indexed in some way, typically based on the values
of its input parameters, so as to facilitate its lookup. The technique of storing
solutions to subproblems instead of recomputing them is called memoization.
2
Dynamic Programming History
Bellman. [1950s] Pioneered the systematic study of dynamic programming.
Etymology.
Dynamic programming = planning over time.
Secretary of Defense was hostile to mathematical research.
Bellman sought an impressive name to avoid confrontation.
"it's impossible to use dynamic in a pejorative sense"

"something not even a Congressman could object to"
Reference: Bellman, R. E. Eye of the Hurricane, An Autobiography.
3
Dynamic Programming Applications
Areas.
Bioinformatics.
Control theory.
Information theory.
Operations research.
Computer science: theory, graphics, AI, compilers, systems, ….
Some famous dynamic programming algorithms.

Unix diff for comparing two files.
Viterbi for hidden Markov models.
Smith-Waterman for genetic sequence alignment.
Bellman-Ford for shortest path routing in networks.
Cocke-Kasami-Younger for parsing context free grammars.
4
6.1 Weighted Interval Scheduling
Weighted Interval Scheduling
Weighted interval scheduling problem.

Job j starts at sj, finishes at fj, and has weight or value vj .
Two jobs compatible if they don't overlap.
Goal: find maximum weight subset of mutually compatible jobs.
h
Time
0 1 2 3 4 5 6 7 8 9 10
6
Unweighted Interval Scheduling Review
Recall. Greedy algorithm works if all weights are 1.

Consider jobs in ascending order of finish time.
Add job to subset if it is compatible with previously chosen jobs.
Observation. Greedy algorithm can fail spectacularly if arbitrary

weights are allowed.
weight = 999 b
weight = 1 a
Time
0 1 2 3 4 5 6 7 8 9 10 11
7
Weighted Interval Scheduling
Notation. Label jobs by finishing time: f1  f2  . . .  fn .

Def. p(j) = largest index i < j such that job i is compatible with j.
Ex: p(8) = 5, p(7) = 3, p(2) = 0.
8
Time
0 1 2 3 4 5 6 7 8 9 10 11
8
Dynamic Programming: Binary Choice
Notation. OPT(j) = value of optimal solution to the problem consisting

of job requests 1, 2, ..., j.
Case 1: OPT selects job j.

– collect profit vj
– can't use incompatible jobs { p(j) + 1, p(j) + 2, ..., j - 1 }
– must include optimal solution to problem consisting of remaining
compatible jobs 1, 2, ..., p(j)
optimal substructure
Case 2: OPT does not select job j.

– must include optimal solution to problem consisting of remaining
compatible jobs 1, 2, ..., j-1
 0 if j  0
OPT( j)  
max  v j  OPT( p( j)), OPT( j 1)  otherwise
9
Weighted Interval Scheduling: Brute Force
Brute force algorithm.
Input: n, s1,…,sn , f1,…,fn , v1,…,vn
Sort jobs by finish times so that f1  f2  ...  fn.
Compute p(1), p(2), …, p(n)
Compute-Opt(j) {
if (j = 0)
return 0
else
return max(vj + Compute-Opt(p(j)), Compute-Opt(j-1))
}
This algorithm can take

exponential time in the
worst-case, due to the
potential redundancy in the
recursive calls.
10
Weighted Interval Scheduling: Brute Force
Observation. Recursive algorithm fails spectacularly because of

redundant sub-problems  exponential algorithms.
Ex. Number of recursive calls for family of "layered" instances grows

like Fibonacci sequence.
1 4 3
2
3 2 2 1
3
4 2 1 1 0 1 0
5
1 0
p(1) = 0, p(j) = j-2
11
Weighted Interval Scheduling: Memoization
Memoization. Store results of each sub-problem in a cache;

lookup as needed.

for j = 1 to n
M[j] = empty
M[0] = 0 global array
M-Compute-Opt(j) {
if (M[j] is empty)
M[j] = max(vj + M-Compute-Opt(p(j)), M-Compute-Opt(j-1))
return M[j]
}
12
Weighted Interval Scheduling: Running Time
Claim. Memoized version of algorithm takes O(n log n) time.

Sort by finish time: O(n log n).
Computing p() : O(n log n) via sorting by start time.
M-Compute-Opt(j): each invocation takes O(1) time and either

– (i) returns an existing value M[j]
– (ii) fills in one new entry M[j] and makes two recursive calls
Progress measure  = # nonempty entries of M[].

– initially  = 0, throughout   n.
– (ii) increases  by 1  at most 2n recursive calls.
Overall running time of M-Compute-Opt(n) is O(n). ▪
Remark. O(n) if jobs are pre-sorted by start and finish times.
13
Weighted Interval Scheduling: Finding a Solution
Q. Dynamic programming algorithms computes optimal value.

What if we want the solution itself?
A. Do some post-processing.
Run M-Compute-Opt(n)
Run Find-Solution(n)
Find-Solution(j) {
if (j = 0)
output nothing
else if (vj + M[p(j)] > M[j-1])
print j
Find-Solution(p(j))
else
Find-Solution(j-1)
}
# of recursive calls  n  O(n).
14
Weighted Interval Scheduling: Bottom-Up
Bottom-up dynamic programming. Unwind recursion.
Iterative-Compute-Opt {
M[0] = 0
for j = 1 to n
M[j] = max(vj + M[p(j)], M[j-1])
}
15
6.3 Segmented Least Squares
Segmented Least Squares
Least squares.
Foundational problem in statistic and numerical analysis.
Given n points in the plane: (x1, y1), (x2, y2) , . . . , (xn, yn).
Find a line y = ax + b that minimizes the sum of the squared error:
n
y
SSE   ( yi  axi  b)2
i1

x
Solution. Calculus  min error is achieved when
n i xi yi  (i xi ) (i yi ) i yi  a i xi
a , b
n i xi  (i xi )
2 2 n
17
Segmented least squares.

Points lie roughly on a sequence of several line segments.
Given n points in the plane (x1, y1), (x2, y2) , . . . , (xn, yn) with
x1 < x2 < ... < xn, find a sequence of lines that minimizes f(x).
Q. What's a reasonable choice for f(x) to balance accuracy and

parsimony?
goodness of fit
number of lines
x
18
Segmented least squares.

Points lie roughly on a sequence of several line segments.
Given n points in the plane (x1, y1), (x2, y2) , . . . , (xn, yn) with
x1 < x2 < ... < xn, find a sequence of lines that minimizes:
– the sum of the sums of the squared errors E in each segment
– the number of lines L
Tradeoff function: E + c L, for some constant c > 0.
x
19
Dynamic Programming: Multiway Choice
Notation.
OPT(j) = minimum cost for points p1, pi+1 , . . . , pj.
e(i, j) = minimum sum of squares for points pi, pi+1 , . . . , pj.
To compute OPT(j):
Last segment uses points pi, pi+1 , . . . , pj for some i.
Cost = e(i, j) + c + OPT(i-1).

 0 if j  0
OPT( j)   min
  e(i, j)  c  OPT(i 1)  otherwise
1  i  j

20
Segmented Least Squares: Algorithm
INPUT: n, p1,…,pN , c
Segmented-Least-Squares() {
M[0] = 0
for j = 1 to n
for i = 1 to j
compute the least square error eij for
the segment pi,…, pj
for j = 1 to n
M[j] = min 1  i  j (eij + c + M[i-1])
return M[n]
}
can be improved to O(n2) by pre-computing various statistics

Running time. O(n3).
Bottleneck = computing e(i, j) for O(n2) pairs, O(n) per pair using
previous formula.
21
6.4 Knapsack Problem
Knapsack Problem
Knapsack problem.
Given n objects and a "knapsack."
Item i weighs wi > 0 kilograms and has value vi > 0.
Knapsack has capacity of W kilograms.
Goal: fill knapsack so as to maximize total value.
# value weight
Ex: { 3, 4 } has value 40.
1 1 1
2 6 2
W = 11
3 18 5
4 22 6
5 28 7
Greedy: repeatedly add item with maximum ratio vi / wi.

Ex: { 5, 2, 1 } achieves only value = 35  greedy not optimal.
23
Dynamic Programming: False Start
Def. OPT(i) = max profit subset of items 1, …, i.
Case 1: OPT does not select item i.

– OPT selects best of { 1, 2, …, i-1 }
Case 2: OPT selects item i.

– accepting item i does not immediately imply that we will have to
reject other items
– without knowing what other items were selected before i,
we don't even know if we have enough room for i
Conclusion. Need more sub-problems!
24
Dynamic Programming: Adding a New Variable
Def. OPT(i, w) = max profit subset of items 1, …, i with weight limit w.
Case 1: OPT does not select item i.

– OPT selects best of { 1, 2, …, i-1 } using weight limit w
Case 2: OPT selects item i.

– new weight limit = w – wi
– OPT selects best of { 1, 2, …, i–1 } using this new weight limit
 0 if i  0

OPT(i, w)  OPT(i 1, w) if w i  w
max OPT(i 1, w), v  OPT(i 1, w  w ) otherwise
  i i 

25
Knapsack Problem: Bottom-Up
Knapsack. Fill up an n-by-W array.
Input: n, W, w1,…,wN, v1,…,vN
for w = 0 to W
M[0, w] = 0
for i = 1 to n
for w = 1 to W
if (wi > w)
M[i, w] = M[i-1, w]
else
M[i, w] = max {M[i-1, w], vi + M[i-1, w-wi ]}
return M[n, W]
26
Knapsack Algorithm
W+1
0 1 2 3 4 5 6 7 8 9 10 11
 0 0 0 0 0 0 0 0 0 0 0 0
{1} 0 1 1 1 1 1 1 1 1 1 1 1
n+1 { 1, 2 } 0 1 6 7 7 7 7 7 7 7 7 7
{ 1, 2, 3 } 0 1 6 7 7 18 19 24 25 25 25 25
{ 1, 2, 3, 4 } 0 1 6 7 7 18 22 24 28 29 29 40
{ 1, 2, 3, 4, 5 } 0 1 6 7 7 18 22 28 29 34 34 40
Item Value Weight

1 1 1
OPT: { 4, 3 } 2 6 2
value = 22 + 18 = 40
W = 11 3 18 5
4 22 6
5 28 7
27
Knapsack Problem: Running Time
Running time. (n W).

Not polynomial in input size!
"Pseudo-polynomial."
Decision version of Knapsack is NP-complete. [Chapter 8]
Knapsack approximation algorithm. There exists a poly-time algorithm

that produces a feasible solution that has value within 0.01% of
optimum. [Section 11.8]
28
Dynamic Programming Summary
Recipe.
Characterize structure of problem.
Recursively define value of optimal solution.
Compute value of optimal solution.
Construct optimal solution from computed information.
Dynamic programming techniques.

Binary choice: weighted interval scheduling.
Viterbi algorithm for HMM also uses
Multi-way choice: segmented least squares. DP to optimize a maximum likelihood
tradeoff between parsimony and accuracy
Adding a new variable: knapsack.
Dynamic programming over intervals: RNA secondary structure.
CKY parsing algorithm for context-free
grammar has similar structure
Top-down vs. bottom-up: different people have different intuitions.
29
6.6 Sequence Alignment
String Similarity
How similar are two strings?

ocurrance o c u r r a n c e -
occurrence
o c c u r r e n c e
6 mismatches, 1 gap
o c - u r r a n c e
o c c u r r e n c e
1 mismatch, 1 gap
o c - u r r - a n c e
o c c u r r e - n c e
0 mismatches, 3 gaps
31
Edit Distance
Applications.
Basis for Unix diff.
Speech recognition.
Computational biology.
Edit distance. [Levenshtein 1966, Needleman-Wunsch 1970]

Gap penalty ; mismatch penalty pq.
Cost = sum of gap and mismatch penalties.
C T G A C C T A C C T - C T G A C C T A C C T
C C T G A C T A C A T C C T G A C - T A C A T
TC + GT + AG+ 2CA 2 + CA
32
Sequence Alignment
Goal: Given two strings X = x1 x2 . . . xm and Y = y1 y2 . . . yn find

alignment of minimum cost.
Def. An alignment M is a set of ordered pairs xi-yj such that each item
occurs in at most one pair and no crossings.
Def. The pair xi-yj and xi'-yj' cross if i < i', but j > j'.
cost(M )    xi y j     
(x i , y j )  M i : x i unmatched j : y j unmatched
mismatch gap
x1 x2 x3 x4 x5 x6
 C T A C C - G
Ex: CTACCG vs. TACATG.
Sol: M = x2-y1, x3-y2, x4-y3, x5-y4, x6-y6.
- T A C A T G
y1 y2 y3 y4 y5 y6
33
Sequence Alignment: Problem Structure
Def. OPT(i, j) = min cost of aligning strings x1 x2 . . . xi and y1 y2 . . . yj.

Case 1: OPT matches xi-yj.
– pay mismatch for xi-yj + min cost of aligning two strings
x1 x2 . . . xi-1 and y1 y2 . . . yj-1
Case 2a: OPT leaves xi unmatched.
– pay gap for xi and min cost of aligning x1 x2 . . . xi-1 and y1 y2 . . . yj
Case 2b: OPT leaves yj unmatched.
– pay gap for yj and min cost of aligning x1 x2 . . . xi and y1 y2 . . . yj-1
 j if i  0
  x i y j  OPT (i 1, j 1)

 
OPT (i, j)   min    OPT (i 1, j) otherwise
    OPT (i, j 1)
 

 i if j  0

34
Sequence Alignment: Algorithm
Sequence-Alignment(m, n, x1x2...xm, y1y2...yn, , ) {

for i = 0 to m
M[i, 0] = i
for j = 0 to n
M[0, j] = j
for i = 1 to m
for j = 1 to n
M[i, j] = min([xi, yj] + M[i-1, j-1],
 + M[i-1, j],
 + M[i, j-1])
return M[m, n]
}
Analysis. (mn) time and space.

English words or sentences: m, n  10.
Computational biology: m = n = 100,000. 10 billions ops OK, but 10GB array?
35
6.7 Sequence Alignment in Linear Space
Sequence Alignment: Linear Space
Q. Can we avoid using quadratic space?
Easy. Optimal value in O(m + n) space and O(mn) time.

Compute OPT(i, •) from OPT(i-1, •).
No longer a simple way to recover alignment itself.
Theorem. [Hirschberg 1975] Optimal alignment in O(m + n) space and

O(mn) time.
Clever combination of divide-and-conquer and dynamic programming.
Inspired by idea of Savitch from complexity theory.
37
Edit distance graph.

Let f(i, j) be shortest path from (0,0) to (i, j).
Observation: f(i, j) = OPT(i, j) (can prove this by induction on i+j)
 y1 y2 y3 y4 y5 y6
 0-0
x1
 xi y j 
x2 
i-j

x3 m-n
38

Let f(i, j) be shortest path from (0,0) to (i, j).
Can compute f (•, j) for any j in O(mn) time and O(m) space.
*Note: we want to recover solution for best alignment in the end; this
will require O(m+n) space.
j
 y1 y2 y3 y4 y5 y6
 0-0
x1
x2 i-j
x3 m-n
39

Let g(i, j) be shortest path from (i, j) to (m, n).
Can compute by reversing the edge orientations and inverting the
roles of (0, 0) and (m, n)
 y1 y2 y3 y4 y5 y6
 0-0

x1 i-j
 xi y j
x2 

x3 m-n
40

Let g(i, j) be shortest path from (i, j) to (m, n).
Can compute g(•, j) for any j in O(mn) time and O(m) space. (just like
we did for f()).
 y1 y2 y3 y4 y5 y6
 0-0
x1 i-j
x2
x3 m-n
41
Observation 1. The cost of the shortest path that uses (i, j) is

f(i, j) + g(i, j).
 y1 y2 y3 y4 y5 y6
 0-0
x1 i-j
x2
x3 m-n
42
Observation 2. let q be an index that minimizes f(q, n/2) + g(q, n/2).

Then, the shortest path from (0, 0) to (m, n) uses (q, n/2).
n/2
 y1 y2 y3 y4 y5 y6
 0-0
x1 i-j q
x2
x3 m-n
43
Divide: find index q that minimizes f(q, n/2) + g(q, n/2) using DP.
Align xq and yn/2.
Conquer: recursively compute optimal alignment in each piece.
n/2
 y1 y2 y3 y4 y5 y6
 0-0
x1 i-j q
x2
x3 m-n
44
Sequence Alignment: Running Time Analysis Warmup
Theorem. Let T(m, n) = max running time of algorithm on strings of

length at most m and n. T(m, n) = O(mn log n).
T(m, n)  2T(m, n/2)  O(mn)  T(m, n)  O(mn log n)

Remark. Analysis is not tight because two sub-problems are of size
(q, n/2) and (m - q, n/2). In next slide, we save log n factor.
45
Sequence Alignment: Running Time Analysis
Theorem. Let T(m, n) = max running time of algorithm on strings of

length m and n. T(m, n) = O(mn).
Pf. (by induction on n)

O(mn) time to compute f( •, n/2) and g ( •, n/2) and find index q.
T(q, n/2) + T(m - q, n/2) time for two recursive calls.
Choose constant c so that:
T(m, 2)  cm
T(2, n)  cn
T(m, n)  cmn  T(q, n /2)  T(m  q, n /2)
Base cases: m = 2 or n = 2.
Inductive hypothesis: T(m, n)
  2cmn.
T (m, n)  T (q, n / 2)  T (m  q, n / 2)  cmn

 2cqn / 2  2c(m  q )n / 2  cmn
 cqn  cmn  cqn  cmn
 2cmn
46
Kleinberg 6.1 (HW)
#1: Let G=(V,E) be an undirected graph with n nodes. Recall that a

subset of the nodes is called an independent set if no two of them are
joined by an edge. Finding large independent sets is difficult in general
(NP-hard); but here we’ll see that it can be done efficiently if the
graph is “simple” enough.
Call a graph G a path if its nodes can be written as v1,v2,…,vn, with an

edge between vi and vj iff the indices differ by one. With each node vi,
we associate a positive integer weight wi.
Goal: Find an independent set in a path G whose total weight is as large

as possible.
47
Kleinberg 6.1 (HW)
#1: Goal: Find an independent set in a path G whose total weight is as

large as possible.
(a) Give an example to show that the following algorithm doesn’t always
find an independent set of max total weight.
Algo: “heaviest-first”
{
S=empty set
while (some node remains in G)
{
Pick node with max weight and add it to S; delete this node and
its neighbors
}
}
48
Kleinberg 6.1 (HW)

large as possible.
(c) Give an algorithm that takes an n-node path G with weights and
returns an independent set of maximum total weight. The running time
should be polynomial in n, independent of the values of the weights.
(Use a DP structure; start with a recursion.)
49
Kleinberg 6.1 (HW)

large as possible.
Let Si denote an independent set on {v1,..,vi}, and let Xi denote its

weight. Define X0=0 and note that X1=w1.
Now for i>1, either vi belong to Si or it doesn’t. What is a natural

recursion?
50
Kleinberg 6.1 (HW)
large as possible.

Now for i>1, either vi belongs to Si or it doesn’t. What is a natural

recursion?
If vi belongs to Si, then vi-1 doesn’t (why?); so either Xi=wi+Xi-2 or Xi=Xi-1

(why?).
51
Kleinberg 6.1 (HW)
large as possible.

Now for i>1, either vi belongs to Si or it doesn’t. What is a natural

recursion?
If vi belongs to Si, then vi-1 doesn’t (why?); so either Xi=wi+Xi-2 or Xi=Xi-1

(why?).
Hence, Xi=max(Xi-1, wi+Xi-2); what’s the run-time to compute Sn?
52
Kleinberg 6.3 (HW)
#3: Let G=(V,E) be a directed graph with nodes v1,…,vn. We say that G
is an ordered graph if it has the following properties:
(i) Each edge goes from a node with a lower index to a node with a
higher index; i.e., each directed edge has the form (vi,vj) with i<j.
(ii) Each node expect vn has at least one edge leaving it. That is, for
every node vi, i=1,2,…,n-1 there is at least one edge of the form
(vi,vj).
The length of a path is the number of edges it contains. The goal is to

solve:
Given an ordered graph G, find the length of the longest path that
begins at v1 and ends at vn.
53
Kleinberg 6.3 (HW)
#3:The length of a path is the number of edges it contains. The goal is
to solve:
Given an ordered graph G, find the length of the longest path that
begins at v1 and ends at vn.
(b) Give an efficient algorithm that takes an ordered graph G and

returns the length of the lonest path that begins at v1 and end at vn.
Idea: Use DP structure; consider subproblems OPT[i], the length of the

longest path from v1 to vi in ordered graph, G. One caveat: not all nodes
vi necessarily have a path from v1 to vi; let’s use the value “-inf” in this
case.
54
Kleinberg 6.3 (HW)
#3:(b) Give an efficient algorithm that takes an ordered graph G and returns the length of
the lonest path that begins at v1 and end at vn.
Idea: Use DP structure; consider subproblems OPT[i], the length of the longest path from
v1 to vi in ordered graph, G. One caveat: not all nodes vi necessarily have a path from v1 to vi;
let’s use the value “-inf” in this case.
Define OPT[1]=0 (base case); use for loop:
M[1]=0
for i=2,..,n
{
M=-inf
for all edges (j,i) in G
{
if M[j]=!-inf && M<M[j]+1
then M=M[j]+1
/endif
} /end for
M[i]=M
} /endfor 55
Kleinberg 6.3 (HW)
#3:(b) Give an efficient algorithm that takes an ordered graph G and

case.
for i=2,..,n
{
M=-inf
{
if M=!-inf && M<M[j]+1 What’s the run time?
then M=M[j]+1
/endif
} /end for
M[i]=M
} /endfor 56
Kleinberg 6.3 (HW)
#3:(b) Give an efficient algorithm that takes an ordered graph G and

case.
for i=2,..,n
{
M=-inf
{
if M=!-inf && M<M[j]+1 What’s the run time? O(n2)
then M=M[j]+1
/endif
} /end for
M[i]=M
} /endfor 57
Kleinberg 6.6 (HW)
58
Kleinberg 6.11 (HW)
#11: Suppose you’re consulting for a company that manufactures PC equipment
and ships it to distributors all over the country. For each of the next n weeks,
they have a projected supply si of equipment (measured in pounds), which has to
be shipped by an air freight carrier.
Each week’s supply can be carried by one of two air freight companies: A or B.
(*) Company A charges a fixed rate r per pound (so it costs r*si to ship a week’s
supply si).
(*) Company B makes contracts for a fixed amount c per week, independent of
the weight. However, contracts with company B must be made in blocks of four
consecutive weeks at a time.
59
Kleinberg 6.11 (HW)
#11: Suppose you’re consulting for a company that manufactures PC equipment
and ships it to distributors all over the country. For each of the next n weeks,
they have a projected supply si of equipment (measured in pounds), which has to
be shipped by an air freight carrier.
Each week’s supply can be carried by one of two air freight companies: A or B.
(*) Company A charges a fixed rate r per pound (so it costs r*si to ship a week’s
supply si).
(*) Company B makes contracts for a fixed amount c per week, independent of
the weight. However, contracts with company B must be made in blocks of four
consecutive weeks at a time.
A schedule, for the PC company, is a choice of air freight company (A or B) for

each of the n weeks, with the restriction that company B, whenever it is chosen,
must be chosen for blocks of four contiguous weeks at a time. The cost of a
schedule is the total amount paid to company A and B.
60
Kleinberg 6.11 (HW)
#11: Give a polynomial-time algorithm that takes a sequence of supply
values s1, s2,….,sn and returns a schedule of minimum cost.
Example. Suppose r=1, c=10, and the sequence of values is:

11,9,9,12,12,12,12,9,9,11.
Then the optimal schedule would be to choose company A for first three weeks,
then company B for a block of four consecutive weeks, and then company A for
the final three weeks.
How to use DP structure to solve?
61
Kleinberg 6.11 (HW)
Let OPT[i] denote the minimum cost of a solution for weeks 1 through i. In an
optimal solution, we either use company A or B for the ith week.
If we use company A, we pay r*si and behave optimally up through week i-1; else
we use company B and pay 4c for this contract, and we behave optimally through
week i-4.
62
Kleinberg 6.11 (HW)
Let OPT[i] denote the minimum cost of a solution for weeks 1 through i. In an
optimal solution, we either use company A or B for the ith week.
If we use company A, we pay r*si and behave optimally up through week i-1; else
we use company B and pay 4c for this contract, and we behave optimally through
week i-4.
In summary, our recursion is as follows: OPT[i]= min(r*si+OPT(i-1), 4c+OPT(i-4)).
What’s the runtime?
63

Algorithm

Uploaded by

Copyright:

Available Formats

Algorithm

Uploaded by

Document Information

Original Description:

Original Title

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

Algorithm

Uploaded by

Copyright:

Available Formats

Chapter 6

CS 350: Winter 2018

Divide-and-conquer. Break up a problem into sub-problems, solve each sub-

Dynamic programming. Break up a problem into a series of overlapping sub-

Bellman. [1950s] Pioneered the systematic study of dynamic programming.

"it's impossible to use dynamic in a pejorative sense"

Reference: Bellman, R. E. Eye of the Hurricane, An Autobiography.

Some famous dynamic programming algorithms.

Weighted interval scheduling problem.

Recall. Greedy algorithm works if all weights are 1.

Observation. Greedy algorithm can fail spectacularly if arbitrary

Notation. Label jobs by finishing time: f1  f2  . . .  fn .

Ex: p(8) = 5, p(7) = 3, p(2) = 0.

Notation. OPT(j) = value of optimal solution to the problem consisting

Case 1: OPT selects job j.

Case 2: OPT does not select job j.

Input: n, s1,…,sn , f1,…,fn , v1,…,vn

Sort jobs by finish times so that f1  f2  ...  fn.

Compute p(1), p(2), …, p(n)

This algorithm can take

Observation. Recursive algorithm fails spectacularly because of

Ex. Number of recursive calls for family of "layered" instances grows

Memoization. Store results of each sub-problem in a cache;

Input: n, s1,…,sn , f1,…,fn , v1,…,vn

Sort jobs by finish times so that f1  f2  ...  fn.

Claim. Memoized version of algorithm takes O(n log n) time.

M-Compute-Opt(j): each invocation takes O(1) time and either

Progress measure  = # nonempty entries of M[].

Overall running time of M-Compute-Opt(n) is O(n). ▪

Remark. O(n) if jobs are pre-sorted by start and finish times.

Q. Dynamic programming algorithms computes optimal value.

# of recursive calls  n  O(n).

Bottom-up dynamic programming. Unwind recursion.

Input: n, s1,…,sn , f1,…,fn , v1,…,vn

Sort jobs by finish times so that f1  f2  ...  fn.

Compute p(1), p(2), …, p(n)

Solution. Calculus  min error is achieved when

Segmented least squares.

Q. What's a reasonable choice for f(x) to balance accuracy and

Segmented least squares.

can be improved to O(n2) by pre-computing various statistics

Greedy: repeatedly add item with maximum ratio vi / wi.

Def. OPT(i) = max profit subset of items 1, …, i.

Case 1: OPT does not select item i.

Case 2: OPT selects item i.

Conclusion. Need more sub-problems!

Def. OPT(i, w) = max profit subset of items 1, …, i with weight limit w.

Case 1: OPT does not select item i.

Case 2: OPT selects item i.

Knapsack. Fill up an n-by-W array.

Input: n, W, w1,…,wN, v1,…,vN

Item Value Weight

Running time. (n W).

Knapsack approximation algorithm. There exists a poly-time algorithm

Dynamic programming techniques.

Top-down vs. bottom-up: different people have different intuitions.

How similar are two strings?

Edit distance. [Levenshtein 1966, Needleman-Wunsch 1970]