Timeline for Is deep learning appropriate to approximate dynamic programming problems?
Current License: CC BY-SA 4.0
3 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jul 29, 2018 at 23:45 | comment | added | Pseudonym♦ | As a general comment, even if $f(x_i \ldots x_j)$ is related in a complex way to $f(x_i \ldots x_{j+1})$, the fact that they are related suggests that there is some kind of locality. So without knowing anything about the problem, I wouldn't rule out the more well-understood class of Monte Carlo algorithms (e.g. Metropolis-Hastings, simulated annealing, evolutionary algorithms, etc). | |
| Jul 29, 2018 at 21:37 | answer | added | D.W.♦ | timeline score: 1 | |
| Jul 29, 2018 at 20:45 | history | asked | lightning | CC BY-SA 4.0 |