Loading...
Please wait, while we are loading the content...
Almost sure convergence of stochastic gradient processes with matrix step sizes
| Content Provider | Hyper Articles en Ligne (HAL) |
|---|---|
| Author | Monnez, Jean-Marie |
| Abstract | We consider a stochastic gradient process, which is a special case of stochastic approximation process, where the positive real step size a_{n} is replaced by a random matrix A_{n}: X_{n+1}=X_{n}-A_{n}∇g(X_{n})-A_{n}V_{n}. We give two theorems of almost sure convergence in the case where the equation ∇g=0 has a set of solutions. |
| Ending Page | 536 |
| Page Count | 6 |
| Starting Page | 531 |
| File Format | |
| ISSN | 01677152 |
| Journal | Statistics and Probability Letters |
| Volume Number | 76 |
| Language | English |
| Publisher | Elsevier |
| Publisher Date | 2006-01-01 |
| Access Restriction | Open |
| Subject Keyword | Stochastic approximation stochastic gradient math Mathematics [math] Statistics [math.ST] |
| Content Type | Text |
| Resource Type | Article |
| Subject | Statistics and Probability Statistics, Probability and Uncertainty |
