Divide and Conquer

Tanaji B Patil

Design & Analysis of Algorithms [23DSEU4P02]

Tanaji B Patil Divide and Conquer 23DSEU4P02 1 / 22

Introduction

Divide-and-Conquer is a problem-solving
strategy that breaks a problem into
subproblems.
If subproblems are still large, then
divide-and-conquer is reapplied.
Each subproblem is solved recursively
and then combined to obtain the final
solution.

Tanaji B Patil Divide and Conquer 23DSEU4P02 2 / 22

General Method

1 Problem:Given function to compute on n input i.e. f (n)

2 Divide: Split the problem into k distinct subsets 1 < k ≤ n
3 Conquer: Solve each subproblem recursively.
4 Combine: Merge the results to get the final solution.

Tanaji B Patil Divide and Conquer 23DSEU4P02 3 / 22

Complexity Analysis

Let the problem size be n.

Assume it divides into ‘a’ subproblems, each of size ‘n/b’.
The recurrence relation is given by:
(
T (1), n=1
T (n) =
aT (n/b) + f (n), n>1

f (n) is the time required for dividing and combining the solutions

Tanaji B Patil Divide and Conquer 23DSEU4P02 4 / 22

Complexity Analysis Example

Consider the case where a = 2, b = 2 and f (n) = n

T (n) = 2T (n/2) + n
= 2[2T (n/4) + n/2] + n
= 4T (n/4) + 2n
= 4[2T (n/8) + n/4] + 2n
= 8T (n/8) + 3n
..
.
= 2i T (n/2i ) + i.n
= 2log n T (n/2log n ) + n log n
= nT (1) + n log n
= n log n + 2n

Tanaji B Patil Divide and Conquer 23DSEU4P02 5 / 22

Binary Search: Recursive

Tanaji B Patil Divide and Conquer 23DSEU4P02 6 / 22

Binary Search: Iterative

Tanaji B Patil Divide and Conquer 23DSEU4P02 7 / 22

Binary Search: Comparisons n = 14

Index: [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

Elements: -15 -6 0 7 9 23 54 82 101 112 125 131 142 151

Comparisons: 3 4 2 4 3 4 1 4 3 4 2 4 3 4

Tanaji B Patil Divide and Conquer 23DSEU4P02 8 / 22

Binary Search: Decision Tree for n = 14

Tanaji B Patil Divide and Conquer 23DSEU4P02 9 / 22

Finding Maximum and Minimum

Possible Improvement:

Best Case: (n − 1) (Elements of a are in increasing order)

Worst Case: 2(n − 1) (Elements of a are in decreasing order)
Average Case: 32 (n − 1)
Tanaji B Patil Divide and Conquer 23DSEU4P02 10 / 22
Finding Maximum and Minimum: Recursive

Tanaji B Patil Divide and Conquer 23DSEU4P02 11 / 22

Finding Maximum and Minimum: Recursive ... Cont

Tanaji B Patil Divide and Conquer 23DSEU4P02 12 / 22

Complexity Analysis of MaxMin


T (n/2) + T (n/2) + 2,
 n>2
T (n) = 1, n=2

0, n=1


When n is a power of 2, i.e. n = 2k the recurrence relation is:

T (n) = 2T (n/2) + 2
= 2[2T (n/4) + 2] + 2
= 4T (n/4) + 4 + 2
= 4[2T (n/8) + 2] + 4 + 2
= 8T (n/8) + 8 + 4 + 2
...
k−1
X
= 2k−1 T (2) + 2i
1

= 2k−1 + 2k − 2
= 3n/2 − 2
Tanaji B Patil Divide and Conquer 23DSEU4P02 13 / 22
Finding Maximum and Minimum: Recursive

Index: [1] [2] [3] [4] [5] [6] [7] [8] [9]
Elements: 22 13 -5 -8 15 60 17 31 47

Tanaji B Patil Divide and Conquer 23DSEU4P02 14 / 22

Merge Sort

Tanaji B Patil Divide and Conquer 23DSEU4P02 15 / 22

Merge Sort

Tanaji B Patil Divide and Conquer 23DSEU4P02 16 / 22

Merge Sort

Tanaji B Patil Divide and Conquer 23DSEU4P02 17 / 22

Complexity Analysis of Merge Sort

(
a n = 1, a constant
T (n) =
2 T (n/2) + c n n > 1, c constant
When n is a power of 2, i.e. n = 2k the recurrence relation is:

T (n) = 2T (n/2) + cn
= 2[2T (n/4) + cn/2] + cn
= 4T (n/4) + 2cn
= 4[2T (n/8) + cn/4] + 2cn
= 8T (n/8) + 3cn
...
= 2k T (1) + kcn
= an + cn log n
...
= O(n log n)

Tanaji B Patil Divide and Conquer 23DSEU4P02 18 / 22

Quick Sort

Tanaji B Patil Divide and Conquer 23DSEU4P02 19 / 22

Quick Sort

Tanaji B Patil Divide and Conquer 23DSEU4P02 20 / 22

Conclusion

Divide-and-Conquer is a fundamental algorithm design paradigm.

Used in sorting, searching, matrix multiplication, and many more problems.
Master Theorem provides a systematic way to analyze its complexity.

Tanaji B Patil Divide and Conquer 23DSEU4P02 21 / 22

 Thank you . . .

Tanaji B Patil Divide and Conquer 23DSEU4P02 22 / 22