0% found this document useful (0 votes)
11 views39 pages

Lecture 33

nub

Uploaded by

jifaha1053
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
11 views39 pages

Lecture 33

nub

Uploaded by

jifaha1053
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 39

Algorithms

Dynamic programming
0-1 Knapsack problem

11/05/24 1
Review: Dynamic programming
 DP is a method for solving certain kind of
problems
 DP can be applied when the solution of a
problem includes solutions to subproblems
 We need to find a recursive formula for the
solution
 We can recursively solve subproblems, starting
from the trivial case, and save their solutions in
memory
 In the end we’ll get the solution of the whole
problem
11/05/24 2
Properties of a problem that can be
solved with dynamic programming
 Simple Subproblems
– We should be able to break the original problem to
smaller subproblems that have the same structure
 Optimal Substructure of the problems
– The solution to the problem must be a composition
of subproblem solutions
 Subproblem Overlap
– Optimal subproblems to unrelated problems can
contain subproblems in common

11/05/24 3
Review: Longest Common
Subsequence (LCS)
 Problem: how to find the longest pattern of
characters that is common to two text
strings X and Y
 Dynamic programming algorithm: solve
subproblems until we get the final solution
 Subproblem: first find the LCS of prefixes
of X and Y.
 this problem has optimal substructure: LCS
of two prefixes is always a part of LCS of
bigger strings
11/05/24 4
Review: Longest Common
Subsequence (LCS) continued
 Define Xi, Yj to be prefixes of X and Y of length
i and j; m = |X|, n = |Y|
 We store the length of LCS(Xi, Yj) in c[i,j]
 Trivial cases: LCS(X0 , Yj ) and LCS(Xi, Y0) is
empty (so c[0,j] = c[i,0] = 0 )
 Recursive formula for c[i,j]:
c[i  1, j  1]  1 if x[i ]  y[ j ],
c[i, j ] 
 max(c[i, j  1], c[i  1, j ]) otherwise
c[m,n] is the final solution
11/05/24 5
Review: Longest Common
Subsequence (LCS)
 After we have filled the array c[ ], we can
use this data to find the characters that
constitute the Longest Common
Subsequence

 Algorithm runs in O(m*n), which is much


better than the brute-force algorithm: O(n
2m)

11/05/24 6
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 , bi and W are integer values)
 Problem: How to pack the knapsack to
achieve maximum total value of packed
items?
11/05/24 7
0-1 Knapsack problem:
a picture
Weight Benefit value
Items
wi bi

2 3
This is a knapsack 3 4
Max weight: W = 20 4 5

5 8
W = 20

9 10

11/05/24 8
0-1 Knapsack problem
 Problem, in other words, is to find
max  bi subject to  wi W
iT iT
 The problem is called a “0-1” problem,
because each item must be entirely accepted
or rejected.
 Just another version of this problem is the
“Fractional Knapsack Problem”, where we
can take fractions of items.
11/05/24 9
0-1 Knapsack problem: brute-
force approach
Let’s first solve this problem with a
straightforward algorithm
 Since there are n items, there are 2n possible
combinations of items.
 We go through all combinations and find
the one with the most total value and with
total weight less or equal to W
 Running time will be O(2n)
11/05/24 10
0-1 Knapsack problem: brute-
force approach
 Can we do better?
 Yes, with an algorithm based on dynamic
programming
 We need to carefully identify the
subproblems
Let’s try this:
If items are labeled 1..n, then a subproblem
would be to find an optimal solution for
Sk = {items labeled 1, 2, .. k}
11/05/24 11
Defining a Subproblem
If items are labeled 1..n, then a subproblem
would be to find an optimal solution for Sk
= {items labeled 1, 2, .. k}
 This is a valid subproblem definition.

 The question is: can we describe the final


solution (Sn ) in terms of subproblems (Sk)?
 Unfortunately, we can’t do that.
Explanation follows….
11/05/24 12
Defining a Subproblem
w1 =2 w2 =4 w3 =5 w4 =3 Weight Benefit
b1 =3 b2 =5 b3 =8 b4 =4
Item wi bi
? #
1 2 3
Max weight: W = 20 S4
For S4: 2 3 4
S5
Total weight: 14; 3 4 5
total benefit: 20 4 5 8
5 9 10
w1 =2 w2 =4 w3 =5 w4 =9
b1 =3 b2 =5 b3 =8 b4 =10

For S5:
Solution for S4 is
Total weight: 20 not part of the
11/05/24 total benefit: 26 solution for S5!!! 13
Defining a Subproblem
(continued)
 As we have seen, the solution for S4 is not part
of the solution for S5
 So our definition of a subproblem is flawed
and we need another one!
 Let’s add another parameter: w, which will
represent the exact weight for each subset of
items
 The subproblem then will be to compute
B[k,w]
11/05/24 14
Recursive Formula for
subproblems
 Recursive formula for subproblems:
 B[k  1, w] if wk  w
B[k , w] 
max{B[k  1, w], B[k  1, w  wk ]  bk } else
 It means, that the best subset of Sk that has total
weight w is one of the two:
1) the best subset of Sk-1 that has total weight w,
or
2) the best subset of Sk-1 that has total weight w-wk
plus the item k
11/05/24 15
Recursive Formula
 B[k  1, w] if wk  w
B[k , w] 
max{B[k  1, w], B[k  1, w  wk ]  bk } else
 The best subset of Sk that has the total weight
w, either contains item k or not.
 First case: wk>w. Item k can’t be part of the
solution, since if it was, the total weight
would be > w, which is unacceptable
 Second case: wk <=w. Then the item k can
be in the solution, and we choose the case
with greater value
11/05/24 16
0-1 Knapsack Algorithm
for w = 0 to W
B[0,w] = 0
for i = 0 to n
B[i,0] = 0
for w = 0 to W
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24 else B[i,w] = B[i-1,w] // wi > w 17
Running time
for w = 0 to W O(W)
B[0,w] = 0
for i = 0 to n Repeat n times
B[i,0] = 0
for w = 0 to W O(W)
< the rest of the code >
What is the running time of this
algorithm?
O(n*W)
Remember that the brute-force algorithm
11/05/24
takes O(2 n
) 18
Example
Let’s run our algorithm on the
following data:

n = 4 (# of elements)
W = 5 (max weight)
Elements (weight, benefit):
(2,3), (3,4), (4,5), (5,6)

11/05/24 19
Example (2)
i 0 1 2 3 4
W
0 0
1 0
2 0
3 0
4 0
5 0

for w = 0 to W
B[0,w] = 0

11/05/24 20
Example (3)
i 0 1 2 3 4
W
0 0 0 0 0 0
1 0
2 0
3 0
4 0
5 0

for i = 0 to n
B[i,0] = 0

11/05/24 21
Items:
Example (4) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 i=1 4: (5,6)
2 0 bi=3
3 0 wi=2
4 0 w=1
5 0 w-wi =-1
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 22
Items:
Example (5) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 i=1 4: (5,6)
2 0 3 bi=3
3 0 wi=2
4 0 w=2
5 0 w-wi =0
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 23
Items:
Example (6) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 i=1 4: (5,6)
2 0 3 bi=3
3 0 3 wi=2
4 0 w=3
5 0 w-wi=1
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 24
Items:
Example (7) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 i=1 4: (5,6)
2 0 3 bi=3
3 0 3 wi=2
4 0 3 w=4
5 0 w-wi=2
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 25
Items:
Example (8) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 i=1 4: (5,6)
2 0 3 bi=3
3 0 3 wi=2
4 0 3 w=5
5 0 3 w-wi=2
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 26
Items:
Example (9) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 0 i=2 4: (5,6)
2 0 3 bi=4
3 0 3 wi=3
4 0 3 w=1
5 0 3 w-wi=-2
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 27
Items:
Example (10) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 0 i=2 4: (5,6)
2 0 3 3 bi=4
3 0 3 wi=3
4 0 3 w=2
5 0 3 w-wi=-1
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 28
Items:
Example (11) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 0 i=2 4: (5,6)
2 0 3 3 bi=4
3 0 3 4 wi=3
4 0 3 w=3
5 0 3 w-wi=0
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 29
Items:
Example (12) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 0 i=2 4: (5,6)
2 0 3 3 bi=4
3 0 3 4 wi=3
4 0 3 4 w=4
5 0 3 w-wi=1
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 30
Items:
Example (13) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 0 i=2 4: (5,6)
2 0 3 3 bi=4
3 0 3 4 wi=3
4 0 3 4 w=5
5 0 3 7 w-wi=2
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 31
Items:
Example (14) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 0 0 i=3 4: (5,6)
2 0 3 3 3 bi=5
3 0 3 4 4 wi=4
4 0 3 4 w=1..3
5 0 3 7
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 32
Items:
Example (15) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 0 0 i=3 4: (5,6)
2 0 3 3 3 bi=5
3 0 3 4 4 wi=4
4 0 3 4 5 w=4
5 0 3 7 w- wi=0
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 33
Items:
Example (15) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 0 0 i=3 4: (5,6)
2 0 3 3 3 bi=5
3 0 3 4 4 wi=4
4 0 3 4 5 w=5
5 0 3 7 7 w- wi=1
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 34
Items:
Example (16) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 0 0 0 i=3 4: (5,6)
2 0 3 3 3 3 bi=5
3 0 3 4 4 4 wi=4
4 0 3 4 5 5 w=1..4
5 0 3 7 7
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 35
Items:
Example (17) 1: (2,3)
i 0 1 2 3 4
W 2: (3,4)
0 0 0 0 0 0 3: (4,5)
1 0 0 0 0 0 i=3 4: (5,6)
2 0 3 3 3 3 bi=5
3 0 3 4 4 4 wi=4
4 0 3 4 5 5 w=5
5 0 3 7 7 7
if wi <= w // item i can be part of the solution
if bi + B[i-1,w-wi] > B[i-1,w]
B[i,w] = bi + B[i-1,w- wi]
else
B[i,w] = B[i-1,w]
11/05/24
else B[i,w] = B[i-1,w] // wi > w 36
Comments
 This algorithm only finds the max possible
value that can be carried in the knapsack
 To know the items that make this maximum
value, an addition to this algorithm is
necessary
 Please see LCS algorithm from the previous
lecture for the example how to extract this
data from the table we built

11/05/24 37
Conclusion
 Dynamic programming is a useful technique
of solving certain kind of problems
 When the solution can be recursively
described in terms of partial solutions, we can
store these partial solutions and re-use them
as necessary
 Running time (Dynamic Programming
algorithm vs. naïve algorithm):
– LCS: O(m*n) vs. O(n * 2m)
– 0-1 Knapsack problem: O(W*n) vs. O(2n)
11/05/24 38
The End

11/05/24 39

You might also like