Timeline for How to generate a sparse inverse covariance matrix for sampling multivariate Gaussian vectors?
Current License: CC BY-SA 3.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 4, 2015 at 20:20 | history | edited | gsmafra | CC BY-SA 3.0 |
edited body
|
| Jun 4, 2015 at 20:02 | comment | added | gsmafra | Every positive-definite matrix has a Cholesky decomposition that takes the form LL' where L is lower triangular (IIRC the inverse is also true), so you could sample L and compute a positive-definite matrix from it. If L is sparse then LL' is also sparse (make sure L is less sparse then what you want your final matrix to be) | |
| Jun 4, 2015 at 19:54 | comment | added | user5054 | Could you explain a little more? | |
| Jun 4, 2015 at 19:52 | history | answered | gsmafra | CC BY-SA 3.0 |