1. Dynamic Programming
Lecture 16

2. Dynamic Programming
 Dynamic programming is typically applied to optimization
problems. In such problems there can be many possible
solutions. Each solution has a value, and we wish to find a
solution with the optimal (minimum or maximum) value.

3. Dynamic Programming
 Like divide and conquer, Dynamic Programming solves
problems by combining solutions to sub problems.
 Unlike divide and conquer, sub problems are not
independent.
 Subproblems may share subsubproblems,
 However, solution to one subproblem may not affect the solutions
to other subproblems of the same problem. (More on this later.)
 Dynamic Programming reduces computation by
 Solving subproblems in a bottom-up fashion.
 Storing solution to a subproblem the first time it is solved.
 Looking up the solution when subproblem is encountered again

4. Dynamic Programming
The development of a dynamic-programming algorithm can be
broken into a sequence of four steps.
1.Characterize the structure of an optimal solution.
2.Recursively define the value of an optimal solution.
3.Compute the value of an optimal solution in a bottom-up
fashion.
4.Construct an optimal solution from computed information.

5. Matrix-chain Multiplication
 Suppose we have a sequence or chain A1, A2, …, An of n
matrices to be multiplied
 That is, we want to compute the product A1A2…An
 There are many possible ways (parenthesizations) to compute
the product

6. Matrix-chain Multiplication
 Example: consider the chain A1, A2, A3, A4 of 4 matrices
 Let us compute the product A1A2A3A4
 There are 5 possible ways:
1. (A1(A2(A3A4)))
2. (A1((A2A3)A4))
3. ((A1A2)(A3A4))
4. ((A1(A2A3))A4)
5. (((A1A2)A3)A4)

7. Algorithm to Multiply 2 Matrices
Input: Matrices Ap×q and Bq×r (with dimensions p×q and q×r)
Result: Matrix Cp×r resulting from the product A·B
MATRIX-MULTIPLY(Ap×q , Bq×r)
1. for i ← 1 to p
2. for j ← 1 to r
3. C[i, j] ← 0
4. for k ← 1 to q
5. C[i, j] ← C[i, j] + A[i, k] · B[k, j]
6. return C

8. Matrix-chain Multiplication
 Example: Consider three matrices A10×100, B100×5, and C5×50
 There are 2 ways to parenthesize
 ((AB)C) = D10×5 · C5×50
 AB ⇒ 10·100·5=5,000 scalar multiplications
 DC ⇒ 10·5·50 =2,500 scalar multiplications
 (A(BC)) = A10×100 · E100×50
 BC ⇒ 100·5·50=25,000 scalar multiplications
 AE ⇒ 10·100·50 =50,000 scalar multiplications

9. Matrix-chain Multiplication Problem
 Matrix-chain multiplication problem
 Given a chain A1, A2, …, An of n matrices, where for i=1, 2, …, n,
matrix Ai has dimension pi-1×pi
 Parenthesize the product A1A2…An such that the total number of
scalar multiplications is minimized
 Brute force method of exhaustive search takes time exponential in
n

10. Step 1: The structure of an optimal
parenthesization
 The optimal substructure of this problem is as follows.
 Suppose that the optimal parenthesization of AiAi+1…Aj splits
the product between Ak and Ak+1. Then the parenthesization
of the subchain AiAi+1…Ak within this optimal
parenthesization of AiAi+1…Aj must be an optimal
parenthesization of AiAi+1…Ak.
 A similar observation holds for the parenthesization of the
subchain Ak+1Ak+2…Aj in the optimal parenthesization of
AiAi+1…Aj.

11. Step 2: A recursive solution
 Let m[i, j] be the minimum number of scalar
multiplications needed to compute the matrix
AiAi+1…Aj; the cost of a cheapest way to compute
A1A2…An would thus be m[1, n].
0
 i= j
m[i, j ] = 
min{m[i, k ] + m[k + 1, j ] + pi −1 pk p j } i < j
 i ≤k < j


12. Step 3: Computing the optimal costs
MATRIX-CHAIN-ORDER(p)
1 n←length[p] - 1
2 for i←1 to n
3 do m[i, i]←0
4 for l←2 to n
5 do for i←1 to n - l + 1
6 do j ←i + l-1
7 m[i, j]←∞
8 for k←i to j - 1
9 do q←m[i, k] + m[k + 1, j] +pi-1pkpj
10 if q < m[i, j]
11 then m[i, j]←q
12 s[i, j]←k
13 return m and s
 It’s running time is O(n3).

13. matrix dimension
----------------------
A1 30 X 35
A2 35 X 15
A3 15 X 5
A4 5 X 10
A5 10 X 20
A6 20 X 25

14. Step 4: Constructing an optimal solution
PRINT-OPTIMAL-PARENS(s, i, ,j)
1 if j =i
2 then print “A”,i
3 else print “(”
4 PRINT-OPTIMAL-PARENS(s, i, s[i, j])
5 PRINT-OPTIMAL-PARENS(s, s[i, j] + 1, j)
6 print “)”
 In the above example, the call PRINT-OPTIMAL-
PARENS(s, 1, 6) computes the matrix-chain product
according to the parenthesization
((A1(A2A3))((A4A5)A6)) .

15. Longest common subsequence
 Formally, given a sequence X ={x1, x2, . . . , xm}, another sequence
Z ={z1, z2, . . . , zk} is a subsequence of X if there exists a strictly
increasing sequence i1, i2, . . . , ik of indices of X such that for all j
= 1, 2, . . . , k, we have xij = zj. For example, Z ={B, C, D, B} is a
subsequence of X ={A, B, C, B, D, A, B } with corresponding
index sequence {2, 3, 5, 7}.

16.  Given two sequences X and Y, we say that a sequence Z is a
common subsequence of X and Y if Z is a subsequence of both X
and Y.
 In the longest-common-subsequence problem, we are given two
sequences X = {x1, x2, . . . , xm} and Y ={y1, y2, . . . , yn} and wish to
find a maximum-length common subsequence of X and Y.

17. Step 1: Characterizing a longest common
subsequence
 Let X ={x1, x2, . . . , xm} and Y ={y1, y2, . . . , yn} be
sequences, and let Z ={z1, z2, . . . , zk} be any LCS of
X and Y.
1. If xm = yn, then zk = xm = yn and Zk-l is an LCS of Xm-l
and Yn-l.
2. If xm≠yn, then zk≠xm implies that Z is an LCS of Xm-1
and Y.
3. If xm≠yn, then zk≠yn implies that Z is an LCS of X
and Yn-l.

18. Step 2: A recursive solution to
subproblems
 Let us define c[i, j] to be the length of an LCS of the sequences Xi
and Yj. The optimal substructure of the LCS problem gives the
recursive formula
0 i = 0orj = 0

c[i, j ] = c[i − 1, j − 1] + 1 i, j > 0andxi = y j
max(c[i, j − 1], c[i − 1, j ]) i, j > 0andx ≠ y
 i j

19. Step 3: Computing the length of an LCS
LCS-LENGTH(X,Y)
1 m←length[X]
2 n←length[Y]
3 for i←1 to m
4 do c[i,0]←0
5 for j←0 to n
6 do c[0, j]←0
7 for i←1 to m
8 do for j←1 to n
9 do if xi = yj
10 then c[i, j]←c[i - 1, j -1] + 1
11 b[i, j] ←”↖”
12 else if c[i - 1, j]←c[i, j - 1]
13 then c[i, j]←c[i - 1, j]
14 b[i, j] ←”↑”
15 else c[i, j]←c[i, j -1]
16 b[i, j] ←”←”
17 return c and b

20. The running time of the procedure is
O(mn).
x=ABCBDAB and y=BDCABA,

21. Step 4: Constructing an LCS
PRINT-LCS(b,X,i,j)
1 if i = 0 or j = 0
2 then return
3 if b[i, j] = “↖"
4 then PRINT-LCS(b,X,i - 1, j - 1)
5 print xi
6 elseif b[i,j] = “↑"
7 then PRINT-LCS(b,X,i - 1,j)
8 else PRINT-LCS(b,X,i, j - 1)

22. Elements of dynamic programming
 For dynamic programming, Optimization problem must
have following to be applicable
 optimal substructure
 overlapping subproblems
 Memoization.

23. Optimal substructure
 The first step in solving an optimization problem by
dynamic programming is to characterize the structure of an
optimal solution. We say that a problem exhibits optimal
substructure if an optimal solution to the problem contains
within it optimal solutions to subproblems. Whenever a
problem exhibits optimal substructure, it is a good clue that
dynamic programming might apply.

24. Overlapping subproblems
 When a recursive algorithm revisits the same problem over
and over again, we say that the optimization problem has
overlapping subproblems

25. Memoization
 A memoized recursive algorithm maintains an entry in a
table for the solution to each subproblem.
 Each table entry initially contains a special value to indicate
that the entry has yet to be filled in.
 When the subproblem is first encountered during the
execution of the recursive algorithm, its solution is
computed and then stored in the table.
 Each subsequent time that the subproblem is encountered,
the value stored in the table is simply looked up and
returned.