用Java实现 Coppersmith Winograd 算法
Coppersmith Winograd 算法是迄今为止已知最快的矩阵乘法算法。具有优先于Strassen计算的渐近运行时间的算法在实践中很少被使用,理由是其运行场合中巨大的稳定因素使它们不合适。
它可以在 O(n^{2.375477}) 时间内将两个 n × sn 矩阵相乘。它用于检查矩阵乘法。
它决定矩阵是否等于选定的 k 值,并且在 O(kn^2) 中,失败的期望值低于 2^-k。
例子
Input:M1={{1,2},
{3,4}}
M2={{3,2},
{5,1}}
Result={{13,4},
{29,10}}
Output:Resultant matrix is matching
算法
// Task is to verify matrix multiplication as M1*M2=M3 or not.
1. Start
2. Take Matrices M1, M2, M3 as an input of (n*n).
3. Choose matrix a[n][1] randomly to which component will be 0 or 1.
4. Calculate M2 * a, M3 * a and then M1 * (M2 * a) for computing the expression,
M1 * (M2 * a) - M3 * a.
5. Verify if M1 * (M2 * a) - M3 * a = 0 or not.
6. If it is zero or false, then matrix multiplication is correct otherwise not.
7. End
下面是上述方法的实现。
Java
// Implementing Coppersmith Winograd Algorithm in Java
import java.io.*;
import java.util.Random;
class GFG {
public static boolean coppersmithWinograd(double[][] M1,
double[][] M2,
double[][] M3, int n)
{
double[][] a = new double[n][1];
Random rand = new Random();
for (int i = 0; i < n; i++) {
a[i][0] = rand.nextInt() % 2;
}
double[][] M2a = new double[n][1];
for (int i = 0; i < n; i++) {
for (int j = 0; j < 1; j++) {
for (int k = 0; k < n; k++) {
M2a[i][j]
= M2a[i][j] + M2[i][k] * a[k][j];
}
}
}
double[][] M3a = new double[n][1];
for (int i = 0; i < n; i++) {
for (int j = 0; j < 1; j++) {
for (int k = 0; k < n; k++) {
M3a[i][j]
= M3a[i][j] + M3[i][k] * a[k][j];
}
}
}
double[][] M12a = new double[n][1];
for (int i = 0; i < n; i++) {
for (int j = 0; j < 1; j++) {
for (int k = 0; k < n; k++) {
M12a[i][j]
= M12a[i][j] + M1[i][k] * M2a[k][j];
}
}
}
for (int i = 0; i < n; i++) {
M12a[i][0] -= M3a[i][0];
}
boolean sameResultantMatrix = true;
for (int i = 0; i < n; i++) {
if (M12a[i][0] == 0)
continue;
else
sameResultantMatrix = false;
}
return sameResultantMatrix;
}
// Driver's Function
public static void main(String[] args)
{
/// "Input the dimension of the matrices: "
int n;
n = 2;
// "Input the 1st or M1 matrix: "
double[][] M1 = { { 1, 2 }, { 3, 4 } };
// "Input the 2nd or M2 matrix: "
double[][] M2 = { { 2, 0 }, { 1, 2 } };
// "Input the result or M3 matrix: "
double[][] M3 = { { 4, 4 }, { 10, 8 } };
if (coppersmithWinograd(M1, M2, M3, n))
System.out.println("Resultant matrix is Matching");
else
System.out.println("Resultant matrix is not Matching");
}
}
输出
Resultant matrix is Matching
时间复杂度: O(n^{2.375477})