INTERNATIONAL JOURNAL OF INFORMATION AND COMPUTING SCIENCE ISSN NO: 0972-1347
A New Approach to Solve Knapsack Problem
[Link]
Department of Computer Science and Engineering, Sant Gadge Baba Amravati University
ankita.pande2@[Link]
[Link]
Department of Computer Science and Engineering, Sant Gadge Baba Amravati University
yugathakare@[Link]
Abstract— The Knapsack Problem belongs to a large class of problems known as Optimization Problem. This problem is to maximize
the obtained profit without exceeding the knapsack capacity. It is a very special case of the well-known Integer Linear Programming
Problem. The purpose of this paper is to analyze several feasible solutions to a Fractional Knapsack Problem using greedy approach.
Based on the knapsack algorithm to take different feasible solutions, in this set of feasible solutions, particular solution that satisfies
the objective of the function. Such solution is called optimum solution. The optimum selection is without revising previously generated
solutions. The greedy choices are made one after the other, reducing each given problem instance to smaller one. The greedy choices
bring efficiency in solving the problem with the help of sub problems.
Keywords- Knapsack, greedy approach,feasible solution,optimal solution,optimization problem.
I. INTRODUCTION TO GREEDY APPROACH
There are several categories of algorithms such as greedy approach, dynamic programming, backtracking, branch and bound
and so on. Knapsack problem is kind of algorithm which falls in greedy category of algorithms. Greedy approach is kind of an
algorithm used for optimization problem. The algorithm which is greedy in nature tries to find a solution which is suitable at
that moment. Greedy approach does not always guarantee optimal solution but when it is used it will gives you most simplest
and efficient solution. Let‟s consider real life example of greedy approach, if you have to travel from source A to source B. And
there are numerous ways to reach to the destination B then in that case greedy method will choose an option that will be best or
suitable at that moment. But it is not necessary that you will definitely reach to destination [Link] method is a straight
forward method. This method is popular for obtaining the optimized solutions. In a greedy technique, the solution is
constructed through a sequence of steps, each step examining a partially constructed solution obtained so far, until a
complete solution to the problem is reached. At each step the choice made should be
Feasible – It should satisfy the problem‟s constraints.
Locally optimal – Among all feasible solutions the best choice is to be made.
Irrevocable – Once the particular choice is made then it should not get changed on subsequent steps.
Volume 6, Issue 1, January 2019 1 [Link]
INTERNATIONAL JOURNAL OF INFORMATION AND COMPUTING SCIENCE ISSN NO: 0972-1347
Actually there are two types of solutions produced by greedy method. One is feasible solution and another one is
optimal solution. Feasible solution is the one which is generalized solution. That means feasible solution is not exact or not
more appropriate solution to the given problem.
For instance consider a quadratic equation
x2 + 5x + 6 = 0……………………………………………………………………………………………………….Equation(1)
The problem is to find out maximum root of the given equation. Then according to the feasible solution answer for equation
(1) would be x = 3 and x = 2. The next category of solutions is optimal solution. Optimal solution is one which gives more exact
or accurate solution to the given problem. Optimal solution for equation (1) would be x=3.
II. KNAPSACK PROBLEM WITH EXAMPLE
The Knapsack Problem is an example of a optimization problem, which seeks for a best solution from among many other
solutions. A thief robbing a store and can carry a maximum weight of Wi into their [Link] are n objects, from
i=1,2,3…n. Each object has a particular weight Wi and obtains a particular profit [Link] object should thief take? This
version of problem is known as Fractional knapsack problem. The setup is same,but the thief can take fractions of items,
meaning that the items can be broken into smaller pieces so that thief may decide to carry only a fraction of xi of item i, where 0
≤ xi ≤ 1. The aim is to fill the knapsack using various items so that the total weight of the items does not exceed the
capacity of the knapsack. To choose only those objects that give maximum profit of the included objects. The total weight of
selected objects should be ≤ M.
For a knapsack problem we have given a sack with „n‟ number of data items has to be filled in that sack. Every data
item has associated its weight value. We have to fill that bag/sack with given data items up to its maximum capacity of sack.
And if that data item is sell out in the market will get its profit value. The main objective of this algorithm is to achieve
maximum profit value in any situation. As this problem is kind of a greedy method then it also has two solutions. These two
solutions with its formulas are as follows.
Feasible Solution
∑ Wi * xi
<=M………………………………………………………………………………………………….……Equation(2)
Optimal Solution
Total Profit =∑ Pi * xi………………………………….……………………………………………………………
Equation(3)
Where, Wi= Weight of an ith object.
Volume 6, Issue 1, January 2019 2 [Link]
INTERNATIONAL JOURNAL OF INFORMATION AND COMPUTING SCIENCE ISSN NO: 0972-1347
Pi=Profit of an ith object
xi= It is the fraction of object that will be inserted in sack.
M=Maximum capacity of the back
There are four strategies to solve the given problem using greedy method.
1) Random Approach
2) Minimum Weight
3) Maximum Profit
4) Maximum Profit/Weight
1) Random Approach-According to this strategy, select an object randomly and add it to that object into the sack. And then
calculate the total profit value of selected object.
2) Minimum Weight-According to this strategy, select an object that is having minimum weight. And then calculate the
total profit value of selected object.
3) Maximum Profit- According to this strategy selects an object that is having maximum weight. And then calculate the
total profit value of selected object.
4) Maximum Profit/Weight- According to this strategy, select an object that is having maximum ratio of Profit/Weight.
And then calculate the total profit value of selected object.
Example 1:-
Consider an example to find an optimal solution to the following instance of knapsack problem if n=5 M=100
(p1,p2,....p5)=(20,30,66,40,60) and (w1,w2,...w5)=(10,20,30,40,50)
Given:-
N 1 2 3 4 5
W 10 20 30 40 50
P 20 30 66 40 60
Volume 6, Issue 1, January 2019 3 [Link]
INTERNATIONAL JOURNAL OF INFORMATION AND COMPUTING SCIENCE ISSN NO: 0972-1347
1) Random Approach
In this strategy, select any of the data items as per your choice randomly. Only make sure that maximum capacity of
sack should not exceed the overall weight of the sack.
N 1 3 4 2 5
W 10 30 40 20 50
P 20 66 40 30 60
Random Approach 1 1 1 1 0
Feasible Solution
∑ Wi * xi <=M
10*1+30*1+40*1+20*1<=100
100<=100 // Condition satisfied
Optimal Solution
Total Profit =∑ Pi * xi
Total Profit =20*1+66*1+40*1+30*1
Total Profit =156
2) Minimum Weight:- In this strategy arrange the given data in ascending order of their weights. And each time check the
weight of the selected object and current capacity of sack so that it should not exceed the maximum capacity of sack.
N 1 2 3 4 5
W 10 20 30 40 50
P 20 30 66 40 60
Minimum Weight 1 1 1 1 0
Volume 6, Issue 1, January 2019 4 [Link]
INTERNATIONAL JOURNAL OF INFORMATION AND COMPUTING SCIENCE ISSN NO: 0972-1347
Feasible Solution
∑ Wi * xi <=M
10*1+20*1+30*1+40*1<=100
100<=100 // Condition satisfied
Optimal Solution
Total Profit =∑ Pi * xi
Total Profit=20*1+30*1+66*1+40*1
Total Profit=156
3) Maximum Profit:-In this strategy arrange the given data in descending order of their profits. And each time check the
weight of the selected object and current capacity of sack so that it should not exceed the maximum capacity of sack.
N 3 5 4 2 1
W 30 50 40 20 10
P 66 60 40 30 20
Maximum Profit 1 1 20/40 0 0
Feasible Solution
∑ Wi * xi <=M
30*1+50*1+40*20/40<=100
100<=100 // Condition satisfied
Optimal Solution
Total Profit =∑ Pi * xi
Volume 6, Issue 1, January 2019 5 [Link]
INTERNATIONAL JOURNAL OF INFORMATION AND COMPUTING SCIENCE ISSN NO: 0972-1347
Total Profit =66*1+60*1+40*20/40+30*0+20*0
Total Profit =146
4) Maximum Profit/Weight:- This is suggested new approach of this [Link] this strategy, first step is to calculate the ratio
of Profit/Weight.
N 1 2 3 4 5
W 10 20 30 40 50
P 20 30 66 40 60
Profit/Weight 2 1.5 2.2 1 1.2
Then arrange the given data in descending order of their profits. Each time check the weight of the selected object and
current capacity of sack so that it should not exceed the maximum capacity of sack.
N 3 1 2 5 4
W 30 10 20 50 40
P 66 20 30 60 40
Profit/Weight 2.2 2 1.5 1.2 1
Maximum Profit/Weight 1 1 1 40/50 0
Feasible Solution
∑ Wi * xi <=M
30*1+10*1+20*1+50*40/50<=100
100<=100 // Condition satisfied
Optimal Solution
Total Profit =∑ Pi * xi
Total Profit =66*1+20*1+30*1+60*40/50
Volume 6, Issue 1, January 2019 6 [Link]
INTERNATIONAL JOURNAL OF INFORMATION AND COMPUTING SCIENCE ISSN NO: 0972-1347
Total Profit =164
After analyzing the above four strategies for solving knapsack problem, it is clearly observed that the last strategy i.e.
Maximum Profit/Weight ration strategy gives us an optimal solution with maximum profit value.
III. ALGORITHM AND ITS DESCRIPTION
Algorithm for Knapsack Problem
Step 1: For i1 to n do
Step 2: If (W[i]>M ) then
Step 3: return;
Step 4: else
Step 5: x[i]=1.0;
Step 6: M=M-W[i]
Step 7: Max_profit=P[i] * x[i]
Step 8: End of If Loop
Step 9: End of For Loop
Step 10: If(i<n) then
Step 11:x[i]=M/W[i];
Step 12: Max_profit = Max_profit + (P[i] * x[i]);
Step 13: End of If Loop
Step 14:Print the Max_profit value.
Where Max_proft=Maximum Profit Value
Wi= Weight of an ith object.
Pi=Profit of an ith object
xi= It is the fraction of object that will be inserted in sack.
M=Maximum capacity of the back
Analysis of Knapsack Algorithm
If the provided items are already sorted into a decreasing order of pi/wi then the for loop takes a time in O(n); Therefore, the
total time including the sort is in O(n logn).
Volume 6, Issue 1, January 2019 7 [Link]
INTERNATIONAL JOURNAL OF INFORMATION AND COMPUTING SCIENCE ISSN NO: 0972-1347
IV. CONCLUSION
This paper proposed an new approach of algorithm to solve the knapsack problem. This algorithm originates from a well-
known continued fraction method and runs in polynomial time of the input length. We conjecture that for any fixed n>2, the
knapsack problem with n variables may be solved in polynomial time. The proof seems difficult and generalization of the
method used in this paper doesn‟t seem to work. It is hoped that the modest beginning presented in this paper will draw the
attention of more researchers and will bring about the eventual solution of this problem.
REFERENCES
[1] Horowitz, E and Sahni, S. “Computing partitions with applications to the knapsack problem”, JACM 21, 2(April 1974),
277-292.
[2] Sri Lakshmi Kanagala, “A Study of Analyzing Greedy Approach for Fractional Knapsack Problem”, IJARCCE,Volume 5,
Issue 2, February [Link] : 2278-102.
[3] Sartaj Sahni, “Data Structures, Algorithms, and Applications in C++”, China Machine Press, (2008).
[4] Parag H. Dave and Himanshu [Link] , “Design and Analysis of Algorithms”, Pearson Education India, 2009.
[5] Aho,Hopcroft & Ullman “The Design & Analysis of Computer Algorithms”, Addison-Wesley.
[6] Cormen, T.H, Lierson & Rivest: “ Introduction to Algorithms”, McGraw-Hill.
Volume 6, Issue 1, January 2019 8 [Link]
