- categories: Linear algebra, Algorithm
Definition
Randomized matrix multiplication approximates the product of two large matrices and by sampling columns (or rows) of and rows (or columns) of according to a probability distribution. This reduces the computational cost while maintaining an acceptable approximation error.
The aim is to compute an approximation of the exact product :
Steps in Randomized Matrix Multiplication
-
Input:
Matrices and , and the number of samples . -
Probability Distribution:
Define probabilities for selecting the -th column of and the corresponding -th row of : -
Sampling:
Sample indices from according to (with replacement). -
Rescale:
Form reduced matrices and by selecting and rescaling the sampled columns of and rows of : -
Approximate Product:
Compute the approximation:
Key Properties
-
Reduction in Complexity:
- Exact multiplication costs .
- Randomized multiplication costs , where .
-
Error Bounds: The Frobenius Norm error is bounded with high probability:
for , given enough samples .
-
Trade-Off: Increasing improves the accuracy of but increases computational cost.
-
Unbiased Estimation: The expectation of the approximation is equal to the exact product: