A multilevel Monte Carlo estimator for matrix multiplication
Inspired by recent developments in multilevel Monte Carlo (MLMC) methods and randomized sketching for linear algebra problems, we propose an MLMC estimator for real-time processing of matrix structured random data. Our algorithm is particularly effective in handling high-dimensional inner products a...
| Autori principali: | , |
|---|---|
| Natura: | Journal article |
| Lingua: | English |
| Pubblicazione: |
Society for Industrial & Applied Mathematics
2020
|
| Riassunto: | Inspired by recent developments in multilevel Monte Carlo (MLMC) methods and randomized sketching for linear algebra problems, we propose an MLMC estimator for real-time processing of matrix structured random data. Our algorithm is particularly effective in handling high-dimensional inner products and matrix multiplication, and finds applications in computer vision and large-scale supervised learning. |
|---|