Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.programming.threads > #2633 > unrolled thread
| Started by | Ramine <ramine@1.1> |
|---|---|
| First post | 2014-10-30 09:21 -0700 |
| Last post | 2014-10-30 10:29 -0700 |
| Articles | 2 — 1 participant |
Back to article view | Back to comp.programming.threads
Strassen algorithm Ramine <ramine@1.1> - 2014-10-30 09:21 -0700
Re: Strassen algorithm Ramine <ramine@1.1> - 2014-10-30 10:29 -0700
| From | Ramine <ramine@1.1> |
|---|---|
| Date | 2014-10-30 09:21 -0700 |
| Subject | Strassen algorithm |
| Message-ID | <m2tdvd$m94$2@dont-email.me> |
Hello, I am using the Strassen algorithm for matrix multiplication inside my multiple linear regression solver, so what about its numerical stability ? Read this: "We begin by discussing Strassen-like algorithms, based on recursive partitioning of matrices into the same number of blocks. We prove that all such algorithms are stable" Read the following paper: http://www.cs.cornell.edu/~rdk/papers/matmul.pdf Thank you, Amine Moulay Ramdane.
[toc] | [next] | [standalone]
| From | Ramine <ramine@1.1> |
|---|---|
| Date | 2014-10-30 10:29 -0700 |
| Message-ID | <m2thu8$5hk$1@dont-email.me> |
| In reply to | #2633 |
Hello, I am also using LU factorization to compute the inverse matrix inside my multiple linear regression solver, so what about its numerical stability ? Read this: "The LU factorization is numerically stable in practice, and produces a reasonable growth factor." Read here on 3.1: http://www.netlib.org/lapack/lawnspdf/lawn259.pdf Amine Moulay Ramdane.
[toc] | [prev] | [standalone]
Back to top | Article view | comp.programming.threads
csiph-web