3 ms·
I may be misunderstanding but they quote >“It only works when your matrix is sparse enough,” said Williams. This is still important but only for this subset o
by a_ghoul 6y ago
I may be misunderstanding but they quote
>“It only works when your matrix is sparse enough,” said Williams.
This is still important but only for this subset of problems really.
- the_svd_doctor 6y agoIt’s a big subset though. Lots of linear systems come from PDE discretization and are structurally sparse with a fixed number of non zeros per row.
- a_ghoul 6y agoThat's very true. I especially would trust someone with your username!
- cambalache 6y agoYes, also the matrices that come out from Gaussian Processes
- Jweb_Guru 6y agoIn practice, the majority of the really big matrices I encounter are very sparse, because they're often constructed as a reduction from some other much smaller representation so we can use linear algebra methods.
- jl2718 6y agoWhat is the actual sparsity criterion to achieve this big-O bound? (Edit: found in arxiv - o(nω−1/log(κ(A))) non-zeros). Note also dependence on condition number.