# Backtracking algorithm-01 knapsack problems

Source: Internet
Author: User

Backpack problems:
Given n items (each item only has one) and a backpack. The Weight of item I is wi, the value is pi, and the size of the backpack is w. Q: How should I select an item and load it into a backpack to maximize the total value?
L if an item can be cut when it is loaded into a backpack, it can be loaded into only one part. In this case, the problem is called a backpack problem.
L when loading a backpack, each type of item I has only two options: loading or not loading, neither loading multiple times nor loading only a part. Therefore, this problem is called the 0-1 backpack problem.
To get the optimal solution, we need to find a balance between the increase in efficiency and the consumption of backpack capacity. That is to say, we should always put those objects with the highest benefits in a backpack first.

Backpack problems can be seen as a backtracing:
Each package is a node, with two candidate values 0 and 1. 0 indicates not putting people in the backpack, and 1 indicates putting them in the backpack.

Therefore, the knapsack problem is converted to finding a path that meets the conditions.

Therefore, the solution can be traced back.

Backtracing solves the problem of a backpack:

Method 1:

// Determine whether the node (I, j) is a node in the solution path:
//
// I indicates the I-th test node in the solution path, and j indicates a candidate value for the node.
// A [I] Save the value selected for node I

BOOL TestNode (I, j)
{
Update related parameter values (assuming this candidate value j is selected, so the affected parameter values are updated );

With 0-(I-1) layer to determine whether there is a conflict with the node to traverse

If a conflict exists, FALSE is returned;
If no conflict exists, save the node I value j to the corresponding array. A [I] = J;

Determine whether I is the last layer,
If it is the last layer, a path is successfully found and TRUE is returned;

If it is not the last layer, determine whether the I + 1 layer has the correct node.

BOOL bFlag = FALSE;
FOR (k = 0; k <CANDIDATA_NUM; k ++) // candidate value [0,..., CANDIDATA_NUM-1]
{
If (TestNode (I + 1, k ))
{
Find a solution;
BFlag = True;
}
// No matter whether TestNode (I + 1, k) is successful or failed, the parameters must be restored after exiting.
Restore related parameter values (the candidate value k is revoked, so the affected parameter values must be restored );

}

RETURN bFlag;

}

[Cpp]
Int m, n = 5, x [10] = {0 };
Int w [6] = {0, 2, 6, 5}, v [6] = {0, 6, 3, 5, 4, 6 };
Int c = 10;
Int cw = 0, cv = 0, bestv = 0;

BOOL TestNode (int I, int j) {// The candidate values for item I are 0 and 1

// Update related parameter values
Cw + = w [I] * j;
Cv + = v [I] * j;

// Conflict with the previous item
If (cw> c)
Return FALSE;

// No conflict. Add it to the solution path.
X [I] = j;

// No conflict exists between 0 and I.
// If the last row successfully finds a solution
If (I = n ){

For (I = 1; I <= n; I ++)
Printf ("% d", x [I]);
If (cv> bestv)
Bestv = cv;
Printf ("\ n ");
Return TRUE;

}

// If it is not the last row, the I + 1 row will be judged

BOOL bSuit = FALSE;
For (int k = 0; k <= 1; k ++)
{
// A proper position exists in line I + 1
If (TestNode (I + 1, k ))
BSuit = TRUE;

// Restore related parameters
Cw-= w [I + 1] * k;
Cv-= v [I + 1] * k;
}

Return bSuit;

}

Void Bag ()
{

For (int I = 0; I <= 1; I ++)
{
TestNode (1, I );
}

}

Or do not update or restore the relevant variables, but deduct the relevant variable values based on known information.

[Cpp]
Int m, n = 5, x [10] = {0 };
Int w [6] = {0, 2, 6, 5}, v [6] = {0, 6, 3, 5, 4, 6 };
Int c = 10;
Int cw = 0, cv = 0, bestv = 0;

BOOL TestNode (int I, int j) {// The candidate values for item I are 0 and 1

// Based on known Parameters
Cw = 0;
Cv = 0;
For (int k = 1; k <= I-1; k ++)
{
Cw + = x [k] * w [k];
Cv + = x [k] * v [k];
}

Cw + = w [I] * j;
Cv + = v [I] * j;

// Conflict with the previous item
If (cw> c)
Return FALSE;

// No conflict. Add it to the solution path.
X [I] = j;

// No conflict exists between 0 and I.
// If the last row successfully finds a solution
If (I = n ){

For (I = 1; I <= n; I ++)
Printf ("% d", x [I]);
If (cv> bestv)
Bestv = cv;
Printf ("\ n ");
Return TRUE;

}

// If it is not the last row, the I + 1 row will be judged

BOOL bSuit = FALSE;
For (int k = 0; k <= 1; k ++)
{
// A proper position exists in line I + 1
If (TestNode (I + 1, k ))
BSuit = TRUE;
}

Return bSuit;

}

Void Bag ()
{

For (int I = 0; I <= 1; I ++)
{
TestNode (1, I );
}

}

Method 2:
The algorithm can be simplified based on the particularity of the knapsack problem.

[Cpp]
Int m, n = 5, x [10] = {0 };
Int w [6] = {0, 2, 6, 5}, v [6] = {0, 6, 3, 5, 4, 6 };
Int c = 10;
Int cw = 0, cv = 0, bestv = 0;

Int OK (int k)
{
Int u = 1;
If (cw> c)
U = 0;
Return u;
}

Int f (int k)
{
Int I;
If (k> n)
{
For (I = 1; I <= n; I ++)
Printf ("% d", x [I]);
If (cv> bestv)
Bestv = cv;
Printf ("\ n ");
}
Else
{
X [k] = 1; // judge candidate value 1
Cw + = w [k];
Cv + = v [k];
If (OK (k ))
F (k + 1 );
Cw-= w [k]; // judge the candidate value 0
Cv-= v [k];
X [k] = 0;
If (OK (k ))
F (k + 1 );
}

Return k;
}

Author: shuilan0066

Related Keywords:
Related Article

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

## A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

• #### Sales Support

1 on 1 presale consultation

• #### After-Sales Support

24/7 Technical Support 6 Free Tickets per Quarter Faster Response

• Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.