Dynamic Programming

Longest Common Subsequence

• Problem: Given 2 sequences, X = x1,...,xm and

Y = y1,...,yn, find a common subsequence whose length is maximum.

springtime ncaa tournament basketball

printing north carolina krzyzewski

Subsequence need not be consecutive, but must be

in order.
Other sequence questions

• Edit distance: Given 2 sequences, X = x1,...,xm and Y = y1,...,yn, what is

the minimum number of deletions, insertions, and changes that you
must do to change one to another?
• Protein sequence alignment: Given a score matrix on amino acid pairs,
s(a,b) for a,b{}A,
and 2 amino acid sequences, X = x1,...,xmAm and Y = y1,...,ynAn, find
the alignment with lowest score…
More problems

Optimal BST: Given sequence K = k1 < k2 <··· < kn of n sorted keys, with a
search probability pi for each key ki, build a binary search tree (BST) with
minimum expected search cost.
Matrix chain multiplication: Given a sequence of matrices A1 A2 … An, with
Ai of dimension mini, insert parenthesis to minimize the total number of
scalar multiplications.
Minimum convex decomposition of a polygon,
Hydrogen placement in protein structures, …
Dynamic Programming

• Dynamic Programming is an algorithm design technique for optimization

problems: often minimizing or maximizing.
• Like divide and conquer, DP solves problems by combining solutions to
subproblems.
• Unlike divide and conquer, subproblems 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.)
• DP 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.
• Key: determine structure of optimal solutions
 Steps in Dynamic Programming

1. Characterize structure of an optimal solution.

2. Define value of optimal solution recursively.
3. Compute optimal solution values either top-down with caching or
bottom-up in a table.
4. Construct an optimal solution from computed values.
We’ll study these with the help of examples.
Longest Common Subsequence

• Problem: Given 2 sequences, X = x1,...,xm and

Y = y1,...,yn, find a common subsequence whose length is maximum.

springtime ncaa tournament basketball

printing north carolina snoeyink

Subsequence need not be consecutive, but must be

in order.
Naïve Algorithm

• For every subsequence of X, check whether it’s a subsequence of Y .

• Time: Θ(n2m).
• 2m subsequences of X to check.
• Each subsequence takes Θ(n) time to check:
scan Y for first letter, for second, and so on.
 Optimal Substructure
Theorem
Theorem
LetZZ==zz1, ,. .. .. ., ,zzkbe
Let
1 k beany
anyLCS
LCSofofXXand
andYY. .
1.1.IfIfxxm ==yyn, ,then
thenzzk ==xxm ==yyn and
andZZk-1 isisan
anLCS
LCSofofXXm-1 and
andYYn-1. .
m n k m n k-1 m-1 n-1
2.2.IfIfxxm yyn, ,then either z  x and Z is an LCS of X and Y .
m n then either zkk  xmm and Z is an LCS of Xm-1
m-1 and Y .
3.3. or zzkkyynnand
or andZZisisan
anLCS
LCSofofXXand
andYYn-1..
n-1

Notation:
prefix Xi = x1,...,xi is the first i letters of X.
This says what any longest common subsequence must look like;
do you believe it?
 Optimal Substructure
Theorem
Theorem
LetZZ==zz1, ,. .. .. ., ,zzkbe
Let
1 k beany
anyLCS
LCSofofXXand
andYY. .
1.1.IfIfxxm ==yyn, ,then
thenzzk ==xxm ==yyn and
andZZk-1 isisan
anLCS
LCSofofXXm-1 and
andYYn-1. .
m n k m n k-1 m-1 n-1
2.2.IfIfxxm yyn, ,then either z  x and Z is an LCS of X and Y .
m n then either zkk  xmm and Z is an LCS of Xm-1
m-1 and Y .
3.3. or zzkkyynnand
or andZZisisan
anLCS
LCSofofXXand
andYYn-1..
n-1
Proof: (case 1: xm = yn)
Any sequence Z’ that does not end in xm = yn can be made longer by adding xm = yn to the end.
Therefore,
(1) longest common subsequence (LCS) Z must end in xm = yn.
(2) Zk-1 is a common subsequence of Xm-1 and Yn-1, and
(3) there is no longer CS of Xm-1 and Yn-1, or Z would not be an LCS.
 Optimal Substructure
Theorem
Theorem
LetZZ==zz1, ,. .. .. ., ,zzkbe
Let
1 k beany
anyLCS
LCSofofXXand
andYY. .
1.1.IfIfxxm ==yyn, ,then
thenzzk ==xxm ==yyn and
andZZk-1 isisan
anLCS
LCSofofXXm-1 and
andYYn-1. .
m n k m n k-1 m-1 n-1
2.2.IfIfxxm yyn, ,then either z  x and Z is an LCS of X and Y .
m n then either zkk  xmm and Z is an LCS of Xm-1
m-1 and Y .
3.3. or zzkkyynnand
or andZZisisan
anLCS
LCSofofXXand
andYYn-1..
n-1

Proof: (case 2: xm  yn, and zk  xm)

Since Z does not end in xm,
(1) Z is a common subsequence of Xm-1 and Y, and
(2) there is no longer CS of Xm-1 and Y, or Z would not be an LCS.
Recursive Solution

• Define c[i, j] = length of LCS of Xi and Yj .

• We want c[m,n].

00 ifif ii00or or jj 00,,


cc[[ii,, jj]]cc[[ii11,, jj11]]11 ifif ii,, jj 00and
andxxi i  yyj j,,
max(c[i  1, j ], c[i, j  1]) ifif ii,, jj 00and
max(c[i  1, j ], c[i, j  1]) andxxi i  yyj j..

This gives a recursive algorithm and solves the problem.

But does it solve it well?
Recursive Solution

00 ififempty
emptyor or empty
empty,,

cc[[,,]]cc[[prefix
prefix,,prefix
prefix]]11 end())end(
ifif end( end()),,
max(c[ prefix ,  ], c[ , prefix ])
max(c[ prefix ,  ], c[ , prefix ]) end())end(
ifif end( end())..

c[springtime, printing]

c[springtim, printing] c[springtime, printin]

[springti, printing] [springtim, printin] [springtim, printin] [springtime, printi]

[springt, printing] [springti, printin] [springtim, printi] [springtime, print]

Recursive Solution
00 ififempty
emptyor or empty
empty,,

cc[[,,]]cc[[prefix
prefix,,prefix
prefix]]11 end())end(
ifif end( end()),,
max(c[ prefix ,  ], c[ , prefix ])
max(c[ prefix ,  ], c[ , prefix ]) end())end(
ifif end( end())..

•Keep track of c[] in a table of

nm entries:
•top/down
•bottom/up
 Constructing an LCS
PRINT-LCS
PRINT-LCS(b, (b,X,
X,i,i,j)j)
1.
1. ififii==00ororjj==00
2.
2. then thenreturn
return
3.
3. ififb[i,
b[i,jj]]==““ ””
4.
4. then thenPRINT-LCS(b,
PRINT-LCS(b,X, X,i 1,j
i1, j1)
1)
5.
5. print
printxxi i
6.
6. elseifelseifb[i,
b[i,jj]]==“↑”“↑”
7.
7. then
thenPRINT-LCS(b,
PRINT-LCS(b,X, X,i
i1,
1,j)j)
8.
8. else
elsePRINT-LCS(b,
PRINT-LCS(b,X, X,i,i,j
j1)
1)
•Initial call is PRINT-LCS (b, X,m, n).
•When b[i, j ] = , we have extended LCS by one character. So LCS = entries with in
them.
•Time: O(m+n)
 Steps in Dynamic Programming
1. Characterize structure of an optimal solution.
2. Define value of optimal solution recursively.
3. Compute optimal solution values either top-down with caching or
bottom-up in a table.
4. Construct an optimal solution from computed values.
We’ll study these with the help of examples.
Optimal Binary Search Trees

• Problem
• Given sequence K = k1 < k2 <··· < kn of n sorted keys,
with a search probability pi for each key ki.
• Want to build a binary search tree (BST)
with minimum expected search cost.
• Actual cost = # of items examined.
• For key ki, cost = depthT(ki)+1, where depthT(ki) = depth of
ki in BST T .
Expected Search Cost

E[search cost in T ]
n
  (depth T (ki )  1)  pi
i 1
n n
  depth T (ki )  pi   pi
i 1 i 1
n Sum of probabilities is 1.
 1   depth T (ki )  pi (15.16)
i 1
 Example

• Consider 5 keys with these search probabilities:

p1k= 0.25, p2 = 0.2, p3 = 0.05, p4 = 0.2, p5 = 0.3.
2

i depthT(ki) depthT(ki)·pi
1 1 0.25
k1 k4 2 0 0
3 2 0.1
4 1 0.2
5 2 0.6
k3 k5
1.15
Therefore, E[search cost] = 2.15.
 Example

• p1 = 0.25, p2 = 0.2, p3 = 0.05, p4 = 0.2, p5 = 0.3.

k2

i depthT(ki) depthT(ki)·pi
1 1 0.25
k1 k5
2 0 0
3 3 0.15
4 2 0.4
k4 5 1 0.3
1.10
Therefore, E[search cost] = 2.10.

k3
This tree turns out to be optimal for this set of keys.
Example

• Observations:
• Optimal BST may not have smallest height.
• Optimal BST may not have highest-probability key at root.
• Build by exhaustive checking?
• Construct each n-node BST.
• For each,
assign keys and compute expected search cost.
• But there are (4n/n3/2) different BSTs with n nodes.
Optimal Substructure

• Any subtree of a BST contains keys in a contiguous range ki, ..., kj for
some 1 ≤ i ≤ j ≤ n.
T

T

• If T is an optimal BST and

T contains subtree T with keys ki, ... ,kj ,
then T must be an optimal BST for keys ki, ..., kj.
• Proof: Cut and paste.
Optimal Substructure

• One of the keys in ki, …,kj, say kr, where i ≤ r ≤ j,

must be the root of an optimal subtree for these k r

keys.
• Left subtree of kr contains ki,...,kr1.
• Right subtree of kr contains kr+1, ...,kkj. k k kj
i r-1 r+1

• To find an optimal BST:

• Examine all candidate roots kr , for i ≤ r ≤ j
• Determine all optimal BSTs containing ki,...,kr1 and containing kr+1,...,kj
Recursive Solution

• Find optimal BST for ki,...,kj, where i ≥ 1, j ≤ n, j ≥ i1.

When j = i1, the tree is empty.
• Define e[i, j ] = expected search cost of optimal BST for
ki,...,kj.

• If j = i1, then e[i, j ] = 0.

• If j ≥ i,
• Select a root kr, for some i ≤ r ≤ j .
• Recursively make an optimal BSTs

• for ki,..,kr1 as the left subtree, and

• for k ,..,kj as the right subtree.
r+1
Recursive Solution

• When the OPT subtree becomes a subtree of a node:

• Depth of every node in OPT subtree goes up by 1.
• Expected search cost increases by
jj
ww((ii,, jj))
ppl l
l l i i from (15.16)

• If kr is the root of an optimal BST for ki,..,kj :

• e[i, j ] = pr + (e[i, r1] + w(i, r1))+(e[r+1, j] + w(r+1, j))
= e[i, r1] + e[r+1, j] + w(i, j).
(because w(i, j)=w(i,r1) + pr + w(r + 1, j))
• But, we don’t know kr. Hence,
00 ifif jjii11
ee[[ii, ,jj]]
min {e[i, r  1]  e[r  1, j ]  w(i, j )} ififii jj
r  j {e[i , r  1]  e[ r  1, j ]  w(i , j )}
i min
 ir  j
Computing an Optimal Solution

For each subproblem (i,j), store:

• expected search cost in a table e[1 ..n+1 , 0 ..n]
• Will use only entries e[i, j ], where j ≥ i1.
• root[i, j ] = root of subtree with keys ki,..,kj, for 1 ≤ i ≤ j
≤ n.
• w[1..n+1, 0..n] = sum of probabilities
• w[i, i1] = 0 for 1 ≤ i ≤ n.
• w[i, j ] = w[i, j-1] + pj for 1 ≤ i ≤ j ≤ n.
Pseudo-code
OPTIMAL-BST(p,
OPTIMAL-BST(p,q,q,n) n)
1.1. for
fori i←
←11to tonn++11 Consider all trees with l keys.
2.2. do e[i,ii1]
doe[i, 1]← ←00 Fix the first key.
3.3. w[i,ii1]
w[i, 1]← ←00 Fix the last key
4.4. for
forl l←
←11to tonn
5.5. do
dofor
fori i←←11to tonnl l++11
6.6. do
doj j←i←i++ll11
7.7. e[i,
e[i,j j]←∞
]←∞
8.8. w[i,
w[i,j j]]← w[i,jj1]
←w[i, 1]++ppj j Determine the root
9.9. for
forrr←i ←ito toj j of the optimal
10.
10. do
dott← e[i,rr1]
←e[i, 1]++e[re[r++1,1,j j]]++w[i,
w[i,j j]] (sub)tree
11.
11. ififtt<<e[i,
e[i,j j]]
12.
12. then
thene[i, e[i,j j]]←
←tt
13.
13. root[i,
root[i,j j]]←r
←r
Elements of Dynamic Programming

• Optimal substructure
• Overlapping subproblems
Optimal Substructure

• Show that a solution to a problem consists of making a choice, which

leaves one or more subproblems to solve.
• Suppose that you are given this last choice that leads to an optimal
solution.
• Given this choice, determine which subproblems arise and how to
characterize the resulting space of subproblems.
• Show that the solutions to the subproblems used within the optimal
solution must themselves be optimal. Usually use cut-and-paste.
• Need to ensure that a wide enough range of choices and subproblems
are considered.
 Optimal Substructure

• Optimal substructure varies across problem domains:

• 1. How many subproblems are used in an optimal solution.
• 2. How many choices in determining which subproblem(s) to use.
• Informally, running time depends on (# of subproblems overall)  (# of
choices).
• How many subproblems and choices do the examples considered
contain?
• Dynamic programming uses optimal substructure bottom up.
• First find optimal solutions to subproblems.
• Then choose which to use in optimal solution to the problem.
Optimal Substucture

• Does optimal substructure apply to all optimization problems? No.

• Applies to determining the shortest path but NOT the longest simple path of an
unweighted directed graph.
• Why?
• Shortest path has independent subproblems.
• Solution to one subproblem does not affect solution to another subproblem
of the same problem.
• Subproblems are not independent in longest simple path.
• Solution to one subproblem affects the solutions to other subproblems.
• Example:
Overlapping Subproblems

• The space of subproblems must be “small”.

• The total number of distinct subproblems is a
polynomial in the input size.
• A recursive algorithm is exponential because it solves the
same problems repeatedly.
• If divide-and-conquer is applicable, then each problem
solved will be brand new.