Lecture 2: Dynamic Programming

1
Lecture 2 Dynamic Programming

• ??????

2
Content

• What is Dynamic Programming?
• Matrix Chain-Products
• Sequence Alignments
• Knapsack Problem
• All-Pairs Shortest Path Problem
• Traveling Salesman Problem
• Conclusion

3
Lecture 2 Dynamic Programming

• What is Dynamic Programming?

4
What is Dynamic Programming?

• Dynamic Programming (DP) tends to break the
  original problem to sub-problems, i.e., in a
  smaller size
• The optimal solution in the bigger sub-problems
  is found through a retroactive formula which
  connects the optimal solutions of sub-problems.
• Used when the solution to a problem may be viewed
  as the result of a sequence of decisions.

5
Properties for Problems Solved by DP

• Simple Subproblems
• The original problem can be broken into smaller
  subproblems with the same structure
• Optimal Substructure of the problems
• The solution to the problem must be a composition
  of subproblem solutions (the principle of
  optimality)
• Subproblem Overlap
• Optimal subproblems to unrelated problems can
  contain subproblems in common

6
The Principle of Optimality

• The basic principle of dynamic programming
• Developed by Richard Bellman
• An optimal path has the property that whatever
  the initial conditions and control variables
  (choices) over some initial period, the control
  (or decision variables) chosen over the remaining
  period must be optimal for the remaining problem,
  with the state resulting from the early decisions
  taken to be the initial condition.

7
Example Shortest Path Problem
Goal
Start
8
Example Shortest Path Problem
Start
Goal
9
Example Shortest Path Problem
25
10
28
5
Start
Goal
40
3
10
Recall ? Greedy Method forShortest Paths on a
Multi-stage Graph
Is the greedy solution optimal?

• Problem
• Find a shortest path from v0 to v3

11
Recall ? Greedy Method forShortest Paths on a
Multi-stage Graph
?
Is the greedy solution optimal?

• Problem
• Find a shortest path from v0 to v3

The optimal path
12
Example ? Dynamic Programming
13
Lecture 2 Dynamic Programming

• Matrix Chain-Products

14
Matrix Multiplication

• C A B
• A is d e and B is e f
• O(def )

15
Matrix Chain-Products

• Given a sequence of matrices, A1, A2, , An, find
  the most efficient way to multiply them together.
• Facts
• A(BC) (AB)C
• Different parenthesizing may need different
  numbers of operation.
• Example A10 30, B 30 5, C 5 60
• (AB)C (10305) (10560) 1500 3000
  4500 ops
• A(BC) (30560) (103060) 9000 18000
  27000 ops

16
Matrix Chain-Products

• Given a sequence of matrices, A1, A2, , An, find
  the most efficient way to multiply them together.
• A Brute-force Approach
• Try all possible ways to parenthesize
  AA1?A2??An
• Calculate number of operations for each one
• Pick the best one
• Time Complexity
• paranethesizations binary trees of n nodes
• O(4n)

17
A Greedy Approach

• Idea 1
• repeatedly select the product that uses the most
  operations.
• Counter-example
• A 10 ? 5, B 5 ? 10, C 10 ? 5, and D 5 ? 10
• Greedy idea 1 gives (AB)(CD), which takes
  5001000500 2000 ops
• A((BC)D) takes 500250250 1000 ops

18
Another Greedy Approach

• Idea 2
• repeatedly select the product that uses the least
  operations.
• Counter-example
• A 101 ? 11, B 11 ? 9, C 9 ? 100, and D 100 ?
  999
• Greedy idea 2 gives A((BC)D), which takes
  1099899900108900228789 ops
• (AB)(CD) takes 99998999189100189090 ops

19
DP ? Define Subproblem
Subproblem (Pij, i? j)
Original Problem
(P1n)
Suppose operations for the optimal solution of
Pij is Nij
operations for the optimal solution of the
original problem P1n is N1n
20
DP ? Define Subproblem
Subproblem (Pij, i? j)
Original Problem
(P1n)
Suppose operations for the optimal solution of
Pij is Nij
operations for the optimal solution of the
original problem P1n is N1n
21
DP ? Define Subproblem
What is the relation btw Nij (Pij) and N1n
(P1n)?
Subproblem (Pij, i? j)
Original Problem
(P1n)
Suppose operations for the optimal solution of
Pij is Nij
operations for the optimal solution of the
original problem P1n is N1n
22
DP ? Principle of Optimality
dk?dj1
di?dk1
Nk1,n
Nik
23
DP ? Implementation
Nij
24
DP ? Implementation
Nij
25
DP ? Implementation
Nij
?
26
DP ? Implementation
Nij
1
2
j
n
1
2
?
i
n
27
DP ? Implementation
Nij
1
2
j
n
1
2
?
i
n
28
DP ? Implementation
Nij
1
2
j
n
1
2
i
n
29
DP ? Implementation
Nij
1
2
j
n
1
?
2
i
n
30
DP ? Implementation
Nij
1
2
j
n
1
2
?
i
n
31
DP ? Implementation
Nij
1
2
j
n
1
2
i
n
32
DP ? Implementation
Nij
1
2
j
n
1
?
2
i
n
33
DP ? Implementation
Nij
1
2
j
n
1
2
i
?
n
34
DP ? Implementation
Nij
1
2
j
n
1
2
i
n
35
DP for Matrix Chain-Products
Algorithm matrixChain(S) Input sequence S of n
matrices to be multiplied Output number of
operations in an optimal parenthesization of
S for i ?1 to n // main diagonal terms are all
zero Ni,i ? 0 for d ? 2 to n // each diagonal
do following for i ?1 to n?d1 // do from top to
bottom for each diagonal j ? id?1 Ni,j ?
infinity for k ? i to j?1 // counting
minimum Ni,j ? min(Ni,j, Ni,k Nk1,j di
dk1 dj1)
36
Time Complexity
Algorithm matrixChain(S) Input sequence S of n
matrices to be multiplied Output number of
operations in an optimal parenthesization of
S for i ?1 to n // main diagonal terms are all
zero Ni,i ? 0 for d ? 2 to n // each diagonal
do following for i ?1 to n?d1 // do from top to
bottom for each diagonal j ? id?1 Ni,j ?
infinity for k ? i to j?1 // counting
minimum Ni,j ? min(Ni,j, Ni,k Nk1,j di
dk1 dj1)
O(n3)
37
Exercises

• The matrixChain algorithm only computes
  operations of an optimal parenthesization. But,
  it doesnt report the optimal parenthesization
  scheme. Please modify the algorithm so that it
  can do so.
• Given an example with 5 matrices to illustrate
  your idea using a table.

38
Lecture 2 Dynamic Programming

• Sequence Alignment

39
Question

• Given two strings
• are they similar?
• what is their distance?

and
40
Example
applicable
X
Y
plausibly
How similar they are?
Can you give them a score?
41
Example
applica---ble
X
Match
Match
Match
Match
Match
Mismatch
Indel
Indel
Indel
Indel
Indel
Indel
Indel
-p-l--ausibly
Y
Matches Mismatches Insertions deletions (indel)
Three cases
42
Example
applica---ble
X
Match
Match
Match
Match
Match
Mismatch
Indel
Indel
Indel
Indel
Indel
Indel
Indel
-p-l--ausibly
Y
Matches Mismatches Insertions deletions (indel)
(1)
(?1)
Three cases
(?1)
43
Example
applica---ble
X
Score 5?(1) 1?(?1) 7 ?(?1) ?3
Match
Match
Match
Match
Match
Mismatch
Indel
Indel
Indel
Indel
Indel
Indel
Indel
-p-l--ausibly
Y
Matches Mismatches Insertions deletions (indel)
(1)
(?1)
Three cases
(?1)
44
Example
applica---ble
X
Is the alignment optimal?
Score 5?(1) 1?(?1) 7 ?(?1) ?3
Match
Match
Match
Match
Match
Mismatch
Indel
Indel
Indel
Indel
Indel
Indel
Indel
-p-l--ausibly
Y
Matches Mismatches Insertions deletions (indel)
(1)
(?1)
Three cases
(?1)
45
Sequence Alignment

• In bioinformatics, a sequence alignment is a way
  of arranging the primary sequences of DNA, RNA,
  or protein to identify regions of similarity that
  may be a consequence of functional, structural,
  or evolutionary relationships between the
  sequences.

46
Global and Local Alignments
L G P S S K Q T G K G S - S R I W D N
Global alignment L N -
I T K S A G K G A I M R L G D A - - - - - - - T
G K G - - - - - - - -
Local alignment - - - - - - - A G K
G - - - - - - - -
47
Global and Local Alignments
48
Global and Local Alignments

• Global Alignment
• attempts to align the entire sequence
• most useful when the sequences in the query set
  are similar and of roughly equal size.
• NeedlemanWunsch algorithm (1971).
• Local Alignment
• Attempts to align partial regions of sequences
  with high level of similarity.
• Smith-Waterman algorithm (1981)

49
NeedlemanWunsch Algorithm

• Find the best global alignment of any two
  sequences under a given substitution matrix.
• Maximize a similarity score, to give maximum
  match
• Maximum match largest number of residues of one
  sequence that can be matched with another
  allowing for all possible gaps
• Based on dynamic programming
• Involves an iterative matrix method of
  calculation

50
Substitution Matrix

• In bioinformatics, a substitution matrix
  estimates the rate at which each possible residue
  in a sequence changes to each other residue over
  time.
• Substitution matrices are usually seen in the
  context of amino acid or DNA sequence alignment,
  where the similarity between sequences depends on
  the mutation rates as represented in the matrix.

51
Substitution Matrix (DNA) w/o Gap Cost
52
Substitution Matrix (DNA) w/ Gap Cost
53
Substitution Matrix (3D-BLAST)
54
DP ? Define Subproblem

• Consider two strings, s of length n and t of
  length m. Let S be the substitution matrix.
• Subproblem Let Pij is defined to be the optimal
  aligning for the two substrings t1..i and
  s1..j,
• and let Mij be the matching score.
• Original Problem Pmn (matching score Mmn)

55
DP ? Principle of Optimality
?
?
?
?
56
Example

• Step 1. Create a scoring matrix
• Step 2. Make an empty table for Mij
• Step 3. Initialize base conditions
• Step 4. Fill table by
• Step 5. Trace back

57
Example

• Step 1. Create a scoring matrix
• Step 2. Make an empty table for Mij
• Step 3. Initialize base conditions
• Step 4. Fill table by
• Step 5. Trace back

0
?2
?4
?6
?8
?10
?12
?14
?2
?4
?6
?8
?10
?12
58
Example

• Step 1. Create a scoring matrix
• Step 2. Make an empty table for Mij
• Step 3. Initialize base conditions
• Step 4. Fill table by
• Step 5. Trace back

0
?2
?4
?6
?8
?10
?12
?14
?2
?1
0
?2
?4
?6
?8
?10
?4
?3
1
?1
?3
?3
?5
?7
?6
?5
?1
0
?2
?1
?3
?5
?8
?4
?3
0
1
?1
1
?1
?10
?6
?5
?2
1
0
1
2
?12
?8
?7
?3
0
0
1
3
59
Example
s t
G G
A A
T -
G A
G T
C C
A C

• Step 1. Create a scoring matrix
• Step 2. Make an empty table for Mij
• Step 3. Initialize base conditions
• Step 4. Fill table by
• Step 5. Trace back

0
?2
?4
?6
?8
?10
?12
?14
0
?2
?1
0
?2
?4
?6
?8
?10
?1
?4
?3
1
?1
?3
?3
?5
?7
1
?6
?5
?1
0
?2
?1
?3
?5
0
?8
?4
?3
0
1
?1
1
?1
?1
1
?10
?6
?5
?2
1
0
1
2
1
?12
?8
?7
?3
0
0
1
3
3
60
NeedlemanWunsch Algorithm

• Step 1. Create a scoring matrix
• Step 2. Make an empty table for Mij
• Step 3. Initialize base conditions
• Step 4. Fill table by
• Step 5. Trace back

61
NeedlemanWunsch Algorithm
s ? , t ? while i lt 1 and j lt 1 do
if s ?sj s t ?ti t else
if s ?sj s t ?gap t
else s ?gap s t ?ti t while
i gt 1 do t ?gap t while j gt 1 do s ?gap s

• Step 1. Create a scoring matrix
• Step 2. Make an empty table for Mij
• Step 3. Initialize base conditions
• Step 4. Fill table by
• Step 5. Trace back

62
Local Alignment Problem

• Given two strings s s1sn,
• t t1.tm
• Find substrings s, t whose similarity
• (optimal global alignment value) is maximum.

63
Example Local Alignment
GTAGT CATCAT ATG TGACTGAC G TC CATDOGCAT CC
TGACTGAC A
Best aligned subsequeces
64
Recursive Formulation

• Global Alignment (NeedlemanWunsch Algorithm)
• Local Alignment (Smith-Waterman Algorithm)

65
Exercises

• Find the best local aligned substrings for the
  following two DNA strings
• GAATTCAGTTA
• GGATCGA
• You have to give the detail.
• Hint start from the left table.

66
Exercises

• What is longest common sequence (LCS) problem?
  How to solve LCS using dynamic programming
  technique?

67
Lecture 2 Dynamic Programming

• Knapsack Problem

68
Knapsack Problems

• Given some items, pack the knapsack to get the
  maximum total value. Each item has some weight
  and some benefit. Total weight that we can carry
  is no more than some fixed capacity.
• Fractional knapsack problem
• Items are divisible you can take any fraction of
  an item.
• Solved with a greedy algorithm.
• 0-1 knapsack problem
• Items are indivisible you either take an item or
  not.
• Solved with dynamic programming.

69
0-1 Knapsack Problem

• Given a knapsack with maximum capacity W, and a
  set S consisting of n items
• Each item i has some weight wi and benefit value
  bi (all wi and W are integer values)
• Problem How to pack the knapsack to achieve
  maximum total value of packed items?

Why it is called a 0-1 Knapsack Problem?
70
Example 0-1 Knapsack Problem
Which boxes should be chosen to maximize the
amount of money while still keeping the overall
weight under 15 kg ?
71
Example 0-1 Knapsack Problem

• Objective Function
• Unknowns or Variables
• Constraints

72
Formulation 0-1 Knapsack Problem
73
0-1 Knapsack Problem Brute-Force Approach

• Since there are n items, there are 2n possible
  combinations of items.
• We go through all combinations and find the one
  with maximum value and with total weight less or
  equal to W
• Running time will be O(2n)

74
DP ? Define Subproblem

• Suppose that items are labeled 1,..., n.
• Define a subproblem, say, Pk as to finding an
  optimal solution for items in Sk 1, 2,..., k.
• ? original problem is Pn.
• Is such a scheme workable?
• Is the principle of optimality held?

75
A Counterexample
1. 2kgs, 3
P1
P2
P3
P4
P5
2. 3kgs, 4
3. 4kgs, 5
4. 5kgs, 8
20 kgs
5. 9kgs, 10
76
A Counterexample
1. 2kgs, 3
2. 3kgs, 4
3. 4kgs, 5
4. 5kgs, 8
20 kgs
5. 9kgs, 10
77
A Counterexample
Solution for P4 is not part of the solution for
P5 !!!
1. 2kgs, 3
2. 3kgs, 4
3. 4kgs, 5
4. 5kgs, 8
20 kgs
5. 9kgs, 10
78
DP ? Define Subproblem

• Suppose that items are labeled 1,..., n.
• Define a subproblem, say, Pk as to finding an
  optimal solution for items in Sk 1, 2,..., k.
• ? original problem is Pn.
• Is such a scheme workable?
• Is the principle of optimality held?

?
?
79
DP ? Define Subproblem
New version

• Suppose that items are labeled 1,..., n.
• Define a subproblem, say, Pk,w as to finding an
  optimal solution for items in Sk 1, 2,..., k
  and with total weight no more than w.
• ? original problem is Pn,W.
• Is such a scheme workable?
• Is the principle of optimality held?

80
DP ? Principle of Optimality
Denote the benefit for the optimal solution of
Pk,w as Bk,w.
81
DP ? Principle of Optimality
In this case, it is impossible to include the kth
object.
Denote the benefit for the optimal solution of
Pk,w as Bk,w.
include the kth object
Not include the kth object
There are two possible choices.
82
Example
1. 2kgs, 3
2. 3kgs, 4
3. 4kgs, 5
4. 5kgs, 6
5 kgs
83
Example
Step 1. Setup table and initialize base
conditions.
1. 2kgs, 3
2. 3kgs, 4
3. 4kgs, 5
4. 5kgs, 6
5 kgs
84
Example
Step 2. Fill all table entries progressively.
1. 2kgs, 3
2. 3kgs, 4
0
3
3. 4kgs, 5
3
4. 5kgs, 6
5 kgs
3
3
85
Example
Step 2. Fill all table entries progressively.
1. 2kgs, 3
2. 3kgs, 4
0
0
3
3
3. 4kgs, 5
3
4
4. 5kgs, 6
5 kgs
3
4
3
7
86
Example
Step 2. Fill all table entries progressively.
1. 2kgs, 3
2. 3kgs, 4
0
0
0
3
3
3
3. 4kgs, 5
3
4
4
4. 5kgs, 6
5 kgs
3
4
5
3
7
7
87
Example
Step 2. Fill all table entries progressively.
1. 2kgs, 3
2. 3kgs, 4
0
0
0
0
3
3
3
3
3. 4kgs, 5
3
4
4
4
4. 5kgs, 6
5 kgs
3
4
5
5
3
7
7
7
88
Example
Step 3. Trace back
1. 2kgs, 3
2. 3kgs, 4
0
0
0
0
3
3
3
3
3. 4kgs, 5
3
4
4
4
4. 5kgs, 6
5 kgs
3
4
5
5
3
7
7
7
89
Pseudo-Polynomial Time Algorithm

• The time complexity for 0-1 knapsack using DP is
  O(Wn).
• Not a polynomial-time algorithm if W is large.
• This is a pseudo-polynomial time algorithm.

90
Lecture 2 Dynamic Programming

• All-Pairs
• Shortest Path Problem

91
All-Pairs Shortest Path Problem

• Given weighted graph G(V,E), we want to determine
  the cost dij of the shortest path between each
  pair of nodes in V.

92
Floyd's Algorithm

• Let be the minimum cost of a path from node i
  to node j, using only nodes in Vkv1,,vk.

k
The all-pairs shortest path problem is to find
all paths with costs
i
j
93
Floyd's Algorithm
Input Parameter D Output Parameter D,
next all_paths(D, next) n D.NumberOfRows //
initialize next if no intermediate // vertices
are allowed nextij j for i 1 to n for j
1 to n nextij j for k 1 to n //
compute D(k) for i 1 to n for j 1 to
n if (Dik Dkj lt Dij)
Dij Dik Dkj nextij
nextik
O(n3)
94
Floyd's Algorithm
Input Parameters next, i, j Output Parameters
None print_path(next, i, j) // if no
intermediate vertices, just // print i and j
and return if (j nextij) print(i
j) return // output i and
then the path from the vertex // after i
(nextij) to j print(i )
print_path(next,nextij, j)
95
Lecture 2 Dynamic Programming

• Traveling Salesman Problem

96
Traveling Salesman Problem (TSP)
97
Traveling Salesman Problem (TSP)
How many feasible paths?
n cities
98
Example (TSP)
(1234) 18
(1243) 19
(1324) 23
(1342) 19
(1423) 23
(1432) 18
99
Subproblem Formulation for TSP
length of the shortest path from i to 1 visiting
each city in S exactly once.
g(i, S)
1
g(1, V ? 1)
length of the optimal TSP tour.
i
100
Subproblem Formulation for TSP
Goal g(1, V ? 1)
length of the shortest path from i to 1 visiting
each city in S exactly once.
g(i, S)
1
j
i
101
Example
Goal g(1, V ? 1)
102
Example
Goal g(1, V ? 1)
18
2
6
4
16
13
14
7
6
7
5
6
5
9
8
13
11
10
9
5
5
6
6
7
7
4
6
4
2
6
2
103
DP ? TSP Algorithm
Goal g(1, V ? 1)
Input Parameter D Output Parameter P //
path TSP(D) n Dim(D) for i 1 to n gi, ?
Di, 1 for k 1 to n?2 // compute g for
subproblems for all S ? V?1 with Sk
for all i ? S ? 1 gi, S minj?SDi,
j, gj, S ? j Pi, S arg minj?SDi,
j, gj, S ? j // compute the TSP tour g1,
V?1 minj?V?1D1, j, gj, V ? 1,
j P1, V?1 arg minj?V?1D1, j, gj,
V ? 1, j
104
DP ? TSP Algorithm
Goal g(1, V ? 1)
Input Parameter D Output Parameter P //
path TSP(D) n Dim(D) for i 1 to n gi, ?
Di, 1 for k 1 to n?2 // compute g for
subproblems for all S ? V?1 with Sk
for all i ? S ? 1 gi, S minj?SDi,
j, gj, S ? j Pi, S arg minj?SDi,
j, gj, S ? j // compute the TSP tour g1,
V?1 minj?V?1D1, j, gj, V ? 1,
j P1, V?1 arg minj?V?1D1, j, gj,
V ? 1, j
O(2n)