DAA Dynamic Programming

Design and Analysis of
Algorithms
Unit - IV
Periyar Govt. Arts College
Cuddalore
Dr. R. Bhuvaneswari
Assistant Professor
Department of Computer Science
Periyar Govt. Arts College, Cuddalore.
1 Dr. R. Bhuvaneswari
Dynamic Programming
General Method:
• It is an algorithm design method that can be used when the solution to a
problem can be viewed as a sequence of decisions.
• It obtains the solution using “Principle of Optimality”.
• It states that “ In an optimal sequence of decisions or choices, each
subsequence must also be optimal”, ie., whatever the initial state and
decision are, the remaining decisions must constitute an optimal
decision sequence.
• The difference between the greedy method and dynamic programming
is that in the greedy method only one decision sequence is ever
generated.
• In dynamic programming, many decision sequences may be generated.
• Sequences containing suboptimal subsequences cannot be optimal and
so will not be generated.
2 Dr. R. Bhuvaneswari Cuddalore
Multistage Graphs
• A multistage graph G = (V, E) is a directed graph in which the vertices

are partitioned into k  2 disjoint sets Vi, 1  i  k.
• If u, v is an edge in E, then u  Vi and v  Vi+1.
• The sets V1 and Vk are such that |V1| = |Vk| = 1.
• The vertex s is the source and the t the sink (destination).
• The multistage graph problem is to find a minimum cost path from s to t.
• The cost of s to t is the sum of the cost of the edges on the path.
• The multistage graph problem can be solved in 2 ways.
Forward method
Backward method

Multistage Graphs
Forward Approach
• In the forward approach, the cost of each and every node is found
starting from the k stage to the 1st stage.
• The minimum cost path from the source to destination is found ie.,
stage 1 to stage k.
• For forward approach,
Cost(i ,j) = min{c(j, l) + cost(i+1, l)}
lVi+1
j, lE
where i is the level number.

Multistage Graphs
V1 V2 V3 V4 V5

Multistage Graphs
Cost(i ,j) = min{c(j, l) + cost(i+1, l)}

lVi+1
j, lE
Min. Cost
cost(5,12) 0 0
cost(4,9) min{c(9,12)+cost(5,12)} = {4 + 0} 4
cost(4,10) min{c(10,12)+cost(5,12)} = {2 + 0} 2
cost(4,11) min{c(11,12)+cost(5,12)} = {5+ 0} 5
cost(3,6) min{c(6,9)+cost(4,9), c(6,10)+cost(4,10)} 7
= min{6+ 4, 5+2}
= min{4+4, 3+2}
= min{5+2, 6+5}

Multistage Graphs
Min. Cost
cost(2,2) min{c(2,6)+cost(3,6), c(2,7)+cost(3,7), 7
c(2,8)+cost(3,8)}
= min{4+7, 2+5, 1+7}
= min{2+7, 7+5}
cost(2,4) min{c(4,8)+cost(3,8)} 18
= min{11+7}
= min{11+5, 8+7}
cost(1,1) min{c(1,2)+cost(2,2), c(1,3)+cost(2,3), 16
c(1,4)+cost(2,4), c(1,5)+cost(2,5)}
= min{9+7, 7+9, 3+18, 2+15}
1  2  7  10  12
1  3  6  10  12 Periyar Govt. Arts College
Multistage Graphs
Algorithm FGraph(G, k, n, p)
//p[1:k] is a minimum cost path
{
cost[n] = 0.0;
for j = n-1 to 1 step -1 do
{
Let r be a vertex such that j, r is an edge of G and c[j, r]+cost[r] is
minimum;
cost[j] = c[j, r] + cost[r];
d[j] = r;
}
p[1] = 1; p[k] = n;
for j = 2 to k-1 do
p[j] = d[p[j-1]];
}
Multistage Graphs
Backward Approach
• In the backward approach, the cost of each and every node is found
starting from the 1st stage to the kth stage.
• The minimum cost path from the source to destination is found ie., stage
k to stage 1.
• For backward approach,
bcost(i, j) = min{bcost(i-1, l) + c(l, j)}
lVi-1
l, jE
where i is the level number.

Multistage Graphs
V1 V2 V3 V4 V5

Multistage Graphs
bcost(i, j) = min{bcost(i-1, l) + c(l, j)}

lVi-1
l, jE
Min. Cost
bcost(1,1) 0 0
bcost(2,2) min{bcost(1,1)+c(1,2)} =min{0+9} 9
bcost(3,6) min{bcost(2,2)+c(2,6),bcost(2,3)+c(3,6)} 9
= min{9+4,7+2}
bcost(3,7) min{bcost(2,2)+c(2,7),bcost(2,3)+c(3,7), 11
bcost(2,5)+c(5,7)}
= min{9+2,7+7,2+11}

Multistage Graphs
Min. Cost
bcost(2,5)+c(5,8)}
= min{9+1,3+11,2+8}
bcost(4,9) min{bcost(3,6)+c(6,9),Bcost(3,7)+c(7,9)} 15
= min{9+6,11+4}
bcost(3,8)+c(8,10)}
= min{9+5,11+3,10+5}
bcost(4,11) min{bcost(3,8)+c(8,11)} = min{10+6} 16
Bcost(5,12) min{bcost(4,9)+c(9,12),bcost(4,10)+c(10,12), 16
bcost(4,11)+c(11,12)}
= min{15+4,14+2,16+5}
12  10  7  2  1 1 2  7  10  12
12  10  6  3  1 1 3  6  10  12 Periyar Govt. Arts College
Multistage Graphs
Algorithm BGraph(G, k, n, p)
{
bcost[1] = 0.0;
for j = 2 to n do
{
Let r be such that r, j is an edge of G and bcost[r] + c[r, j] is minimum;
bcost[j] = bcost[r] + c[r, j];
d[j] = r;
}
p[1] = 1; p[k] = n;
for j = k-1 to 2 step -1 do
p[j] = d[p[j+1]];
}

All pair shortest paths
• All pairs shortest path problem is the determination of the shortest graph
distances between every pair of vertices in a given directed graph G.
• That is, for every pair of vertices (i, j), we are to find a shortest path from
i to j as well as from j to i. These two paths are the same when G is
undirected.
• Let G = (V, E) be a directed graph with n vertices.
• Let cost be a cost adjacency matrix for G such that cost(i, i) = 0, 1  i  n.
• Cost(i, j) is the length of edge i, j if i, j  E(G) and cost(i, j) =  if i  j
and i, j  E(G).
• All pair shortest path problem is to determine a matrix A such that A(i, j)
is the length of a shortest path from i to j.
• Since each application of this procedure requires O(n2) time, the matrix A
can be obtained in O(n3) time.
• The shortest i to j path in G, i ≠ j originates at vertex i and goes through

some intermediate vertices and terminates at vertex j.
• If k is an intermediate vertex on this shortest path, then the subpaths from
i to k and from k to j must be shortest paths from i to k and k to j,
respectively.
• Otherwise, the i to j path is not of minimum length.
• So, the principle of optimality holds.
• Let Ak(i, j) represent the length of a shortest path from i to j going through
no vertex of index greater than k, we obtain:
𝑨𝒌 𝒊, 𝒋 = 𝐦𝐢𝐧 {𝑨𝒌−𝟏 𝒊, 𝒌 + 𝑨𝒌−𝟏 𝒌, 𝒋 , 𝒄𝒐𝒔𝒕 𝒊, 𝒋 }

𝟏 ≤𝒌 ≤𝒏
• Time complexity of this algorithm is O(n3)

Algorithm AllPaths(cost, A, n) 6
{ 4
for i =1 to n do 1 2
{
for j = 1 to n do 11
2
A[i, j] = cost[i, j]; 3
} 3
for k = 1 to n do
{ 1->3 = 11
for j = 1 to n do 1->2->3 = 6
{ 2->1 = 6
for j = 1 to n do 2->3->1 = 5
A[i, j] = min{A[i, j], A[i, k]+A[k, j]};
}
}
} Periyar Govt. Arts College
Solve the problem for k = 1, 2, 3

A0 1 2 3
Cost adjacency matrix
1 0 4 11
2 6 0 2
3 3  0
Solving the Solving the Solving the

equation for, k = 1 equation for, k = 2 equation for, k = 3
A1 1 2 3 A2 1 2 3 A3 1 2 3
1 0 4 11 1 0 4 6 1 0 4 6
2 6 0 2 2 6 0 2 2 5 0 2
3 3 7 0 3 3 7 0 3 3 7 0

String Editing
• Given two strings X = x1, x2, …. , xn and Y = y1, y2, ….., yn, where xi,
1  i  n, and yj, 1  j  m, are members of a finite set of symbols known
as the alphabet.
• We want to transform X into Y using a sequence of edit operations on X.
• The permissible edit operations are insert, delete and change, and there is a
cost associated with each operation.
• The cost of a sequence of operations is the sum of the costs of the
individual operations in the sequence.
• The problem of string editing is to identify a minimum-cost sequence of
edit operations that will transform X into Y.
• D(xi) – cost of deleting the symbol xi from X
• I(yj) – the cost of inserting the symbol yj into X
• C(xi, yj) – cost of changing the symbol xi of X into yj
• Cost of changing any symbol to any other symbol is 2.
String Editing
• Cost associated with each insertion and deletion is 1.

• A dynamic programming solution to this problem can be obtained as
follows.
• Define cost(i, j) be the minimum cost of any edit sequence for
transforming x1, x2, …, xi into y1, y2, …, yj
• Compute cost(i, j) for each i and j.
• Then cost(n, m) is the cost of an optimal edit sequence.
• For i = j = 0, cost(i, j) = 0, since the two sequences are identical and
empty.
• If j = 0 and i > 0, we can transform X into Y by a sequence of deletes.
Cost(i, 0) = cost(i-1, 0) + D(xi).
• If i = 0 and j > 0, we can transform X into Y by a sequence of insertions
Cost(0, j) = cost(0, j-1) + I(yj)
String Editing
• If i  0 and j  0, x1, x2, …., xi can transformed into y1, y2, … yj in one of
the following ways:
 Transform x1, x2, …., xi-1 into y1, y2, …, yj using a minimum-cost edit
sequence and then delete xi. The cost is cost(i-1, j) + D(xi)
 Transform x1, x2, …., xi-1 into y1, y2, …, yj-1 using a minimum-cost edit
sequence and then change the symbol xi to yj. The cost is cost(i-1, j-1)
+ C(xi, yj)
 Transform x1, x2, …., xi into y1, y2, …, yj-1 using a minimum-cost edit
sequence and then insert yj. The cost is cost(i, j-1) + I(yj)
• The minimum cost of any edit sequence is the minimum of the above
three costs, according to the principle of optimality.

String Editing
• The recurrence equation for cost(i, j) is

𝟎 𝒊=𝒋=𝟎
𝒄𝒐𝒔𝒕 𝒊 − 𝟏, 𝟎 + 𝑫 𝒙𝒊 𝒋 = 𝟎, 𝒊 > 0
𝑐𝒐𝒔𝒕 𝒊, 𝒋 =
𝒄𝒐𝒔𝒕 𝟎, 𝒋 − 𝟏 + 𝑰 𝒚𝒋 𝒊 = 𝟎, 𝒋 > 0
𝒄𝒐𝒔𝒕′ 𝒊, 𝒋 𝒊 > 0, 𝒋 > 0
where
cost’ i, j = min{cost i-1, j) +D(xi), cost(i-1, j-1) + C(xi, yj),
cost(i, j-1) + I(yj)}

String Editing
Given two strings, X and Y and edit operations (given below). Convert
string X into Y with minimum number of operations.
Allowed Operations:
 Insertion – Insert a new character.
 Deletion – Delete a character.
 Replace – Replace one character by another.
Example 1:
String X = "horizon"
String Y = "horzon"
Output: {remove 'i' from string X}
Example 2:
String X = a, a, b, a, b
String Y = b, a, b, b
For the cases i = 0, j > 1, and j = 0, i > 1, cost(i, j) can be computed first
and tabulated in the form of a table. The rest of the entries in the table can
be computed in the row-major order.
String Editing
Cost(1,1) = min{cost(0,1)+D(x1), cost(0,0) + C(x1,y1), cost(1,0)+I(y1)}

= min{1+1, 0+2, 1+1} = 2
= min{2+1, 1+0, 2+1} = 1
= min{3+1, 2+2, 1+1} = 2
= min{4+1, 3+2, 2+1} = 3
= min{2+1, 1+2, 2+1} = 3
= min{1+1, 2+0, 3+1} = 2
= min{2+1, 1+2, 2+1} = 3
= min{3+1, 2+2, 3+1} = 4 Periyar Govt. Arts College
String Editing

= min{3+1, 2+0, 3+1} = 2
= min{2+1, 3+2, 2+1} = 3
= min{3+1, 2+0, 3+1} = 2
= min{4+1, 3+0, 2+1} = 3
= min{2+1, 3+2, 4+1} = 3
= min{3+1, 2+0, 3+1} = 2
= min{2+1, 3+2, 2+1} = 3
= min{3+1, 2+2, 3+1} = 4 Periyar Govt. Arts College
String Editing
= min{3+1, 4+0, 5+1} = 4
= min{2+1, 3+2, 4+1} = 3
= min{3+1, 2+0, 3+1} = 2
= min{4+1, 3+0, 2+1} = 3
Optimal operations are:
 Insert y1, delete x2 and x4
 Change x1 by y1, delete x4
 Delete x1 and x2, insert y3
 Delete x1 and x2, insert y4
Time complexity: O(mn)
0/1 Knapsack Problem
• Given n objects and a knapsack or bag.

• wi → weight of object i.
• m → knapsack capacity.
• As the name suggests, objects are indivisible in this method. No
fractional objects can be taken. An object can either be taken completely
or left completely.
• Objective is to fill the knapsack that maximizes the total profit earned.
• Problem can be stated as
maximize pi xi
1 ≤ i ≤n
subject to wi x i ≤ m
1≤ i ≤n
xi = 0 or 1, 1 ≤ i ≤ n
0/1 knapsack problem is solved using dynamic programming in the

following steps-
• Draw a table say ‘V’ with (n+1) number of rows and (w+1) number of
columns.
• Fill all the boxes of 0th row and 0th column with zeroes.
• Start filling the table row wise top to bottom from left to right.
• Use the following formula:
V[i ,W] = max{V[i-1 ,W] , V[i-1, W – w[i]] + p[i]}
• value of the last box represents the maximum possible value that can be
put into the knapsack.

• To identify the items that must be put into the knapsack to obtain the
maximum profit,
 Consider the last column of the table.
 Start scanning the entries from bottom to top.
 On encountering an entry whose value is not same as the value
stored in the entry immediately above it, mark the row label of that
entry.
 After all the entries are scanned, the marked labels represent the
items that must be put into the knapsack.
• O(nw) time is taken to solve 0/1 knapsack problem using dynamic
programming.

Pi = {1, 2, 5, 6}
wi = (2, 3, 4, 5}
m = 8, n = 4
V[i ,W] = max{V[i-1 ,W] , V[i-1, W – w[i]] + p[i]}

V[1,1] = max{V[0,1], V[0,1-2]+1} = max{0, -} = 0
V[1,2] = max{V[0,2], V[0,2-2]+1} = max{0, 0+1} = 1
V[1,3] = max{V[0,3], V[0,3-2]+1} = max{0,0+1} = 1
V[1,4] = max{V[0,4], V[0,4-2]+1} = max{0, 0+1} = 1
V[1,5] = max{V[0,5], V[0,5-2]+1} = max{0, 0+1} = 1
V[1,6] = max{V[0,6], V[0,6-2]+1} = max{0, 0+1} = 1
V[1,7] = max{V[0,7], V[0,7-2]+1} = max{0, 0+1} = 1
V[1,8] = max{V[0,8], V[0,8-2]+1} = max{0, 0+1} = 1
V[2,1] = max{V[1,1],V[1,1-3]+2} = max{0, -} = 0

V[2,2] = max{V[1,2],V[1,2-3]+2} = max{1, -} = 1
V[2,3] = max{V[1,3],V[1,3-3]+2} = max{1, 0+2} = 2
V[2,4] = max{V[1,4],V[1,4-3]+2} = max{1, 0+2} = 2
V[2,5] = max{V[1,5],V[1,5-3]+2} = max{1, 1+2} = 3
V[2,6] = max{V[1,6],V[1,6-3]+2} = max{1, 1+2} = 3
V[2,7] = max{V[1,7],V[1,7-3]+2} = max{1, 1+2} = 3
V[2,8] = max{V[1,8],V[1,8-3]+2} = max{1, 1+2} = 3
V[3,1] = max{V[2,1],V[2,1-4]+5} = max{0, -} = 0

V[3,2] = max{V[2,2],V[2,2-4]+5} = max{1, -} = 1
V[3,3] = max{V[2,3],V[2,3-4]+5} = max{2, -} = 2
V[3,4] = max{V[2,4],V[2,4-4]+5} = max{2, 0+5} = 5
V[3,5] = max{V[2,5],V[2,5-4]+5} = max{2, 0+5} = 5
V[3,6] = max{V[2,6],V[2,6-4]+5} = max{2, 1+5} = 6
V[3,7] = max{V[2,7],V[2,7-4]+5} = max{2, 2+5} = 7

V[3,8] = max{V[2,8],V[2,8-4]+5} = max{2, 2+5} = 7
V[4,1] = max{V[3,1],V[3,1-5]+6} = max{0, -} = 0

V[4,2] = max{V[3,2],V[3,2-5]+6} = max{1, -} = 1
V[4,3] = max{V[3,3],V[3,3-5]+6} = max{2, -} = 2
V[4,4] = max{V[3,4],V[3,4-5]+6} = max{5, -} = 5
V[4,5] = max{V[3,5],V[3,5-5]+6} = max{5, 0+6} = 6
V[4,6] = max{V[3,6],V[3,6-5]+6} = max{6, 0+6} = 6
V[4,7] = max{V[3,7],V[3,7-5]+6} = max{7, 1+6} = 7
V[4,8] = max{V[3,8],V[3,8-5]+6} = max{7, 2+6} = 8
x1 = 0, x2 = 1, x3 = 0, x4 = 1

for (i = 0; i  n; i++)
{
for(w = 0; w  m; w++)
{
if(i==0 || w==0)
k[i][w] = 0;
else if(wt[i]  w)
k[i][w] = max(p[i]+k[i-1][w-wt[i], k[i-1][w]);
else
k[i][w] = k[i-1][w];
}
}

Traveling Salesperson Problem
• The traveling salesperson problem is to find a tour of minimum cost.

• Let G = (V, E) be a directed graph with edge cost Cij =  if i, j  E
• Let V = n and assume n > 1
• A tour G is a directed simple cycle that includes every vertex in V.
• The cost of a tour is the sum of the cost of the edges on the tour.
• Let g(i, S) be the length of a shortest path starting at vertex i, going
through all vertices in S and terminating at vertex 1.
• The function g(1, V-{1}) is the length of an optimal salesperson tour.
𝑔 1, 𝑉 − {1 = min {𝑐1𝑘 + 𝑔 𝑘, 𝑉 − {1, 𝑘 −−−−−1

2 ≤𝑘 ≤𝑛
𝑔 𝑖, 𝑆 = min 𝑐𝑖𝑗 + 𝑔 𝑗, 𝑆 − {𝑗 −−−−−2

𝑗 ∈𝑆

g(i,) = Ci1, 1  i  n.
S = 
g(2,) = c21 = 5
g(3,) = c31 = 6
g(4, ) = c41 = 8
1 2 3 4
Using equation 2, we obtain 1 0 10 15 20
2 5 0 9 10
S = 1 3 6 13 0 12
g(2,{3}) = c23 + g(3,) = 9+6 = 15 4 8 8 9 0
g(2,{4}) = c24 + g(4,) = 10+8 = 18
g(3,{2}) = c32 +g(2,) = 13+5 = 18
g(3,{4}) = c34+g(4,) = 12+8 = 20
g(4,{2}) = c42+g(2,) = 8+5 = 13
g(4,{3}) = c43+g(3,) = 9+6 = 15
S = 2
g(2,{3,4}) = min{c23 +g(3,{4}), c24 + g(4,{3})}
= min{9+20, 10+15} = 25
g(3,{2,4}) = min{c32 + g(2,{4}), c34 + g(4,{2})}
= min{13+18, 12+13} = 25
g(4,{2,3}) = min{c42 + g(2,{3}), c43 + g(3,{2})}
= min{8+15, 9+18} = 23
Using equation 1, we obtain
g(1,{2,3,4}) = min{c12+g(2,{3,4}), c13+g(3,{2,4}), c14+g(4,{2,3})}
= min{10+25, 15+25, 20+23} = 35
The optimal tour is
1->2->4->3->1
O(2nn2) time is taken to solve the traveling salesperson problem


DAA Dynamic Programming

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

DAA Dynamic Programming

Uploaded by

Copyright:

Available Formats

Design and Analysis of

• A multistage graph G = (V, E) is a directed graph in which the vertices

Periyar Govt. Arts College

Periyar Govt. Arts College

Periyar Govt. Arts College

Cost(i ,j) = min{c(j, l) + cost(i+1, l)}

Periyar Govt. Arts College

Periyar Govt. Arts College

Periyar Govt. Arts College

bcost(i, j) = min{bcost(i-1, l) + c(l, j)}

Periyar Govt. Arts College

Periyar Govt. Arts College

• The shortest i to j path in G, i ≠ j originates at vertex i and goes through

𝑨𝒌 𝒊, 𝒋 = 𝐦𝐢𝐧 {𝑨𝒌−𝟏 𝒊, 𝒌 + 𝑨𝒌−𝟏 𝒌, 𝒋 , 𝒄𝒐𝒔𝒕 𝒊, 𝒋 }

• Time complexity of this algorithm is O(n3)

Periyar Govt. Arts College

Solve the problem for k = 1, 2, 3

Solving the Solving the Solving the

Periyar Govt. Arts College

• Cost associated with each insertion and deletion is 1.

Periyar Govt. Arts College

• The recurrence equation for cost(i, j) is

Periyar Govt. Arts College

Cost(1,1) = min{cost(0,1)+D(x1), cost(0,0) + C(x1,y1), cost(1,0)+I(y1)}

Cost(3,1) = min{cost(2,1)+D(x3), cost(2,0) + C(x3,y1), cost(3,0)+I(y1)}

• Given n objects and a knapsack or bag.

0/1 knapsack problem is solved using dynamic programming in the

Periyar Govt. Arts College

Periyar Govt. Arts College

V[i ,W] = max{V[i-1 ,W] , V[i-1, W – w[i]] + p[i]}

V[2,1] = max{V[1,1],V[1,1-3]+2} = max{0, -} = 0

V[3,1] = max{V[2,1],V[2,1-4]+5} = max{0, -} = 0

V[3,7] = max{V[2,7],V[2,7-4]+5} = max{2, 2+5} = 7

V[4,1] = max{V[3,1],V[3,1-5]+6} = max{0, -} = 0

Periyar Govt. Arts College

Periyar Govt. Arts College

• The traveling salesperson problem is to find a tour of minimum cost.

𝑔 1, 𝑉 − {1 = min {𝑐1𝑘 + 𝑔 𝑘, 𝑉 − {1, 𝑘 −−−−−1

𝑔 𝑖, 𝑆 = min 𝑐𝑖𝑗 + 𝑔 𝑗, 𝑆 − {𝑗 −−−−−2

Periyar Govt. Arts College

O(2nn2) time is taken to solve the traveling salesperson problem

You might also like