Timeline for Hyper parameters tuning: Random search vs Bayesian optimization
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 15, 2018 at 20:51 | answer | added | hlsmith | timeline score: 2 | |
| Dec 7, 2017 at 5:52 | comment | added | JPJ | Google is selling their deep learning cloud services now and pushing a feature that automatically tunes your hyperparameters with Bayesian optimization...of course claiming it does the best and is faster as well (searching the hyperspace more efficiently). There are several papers out there that evaluate BO vs RS as well showing BO doing just slightly better. IMO from what I've seen, the diff is something you'd care about more in a Kaggle competition than real life. | |
| Sep 13, 2017 at 8:10 | answer | added | Fabian Werner | timeline score: 16 | |
| Sep 13, 2017 at 8:07 | answer | added | itdxer | timeline score: 10 | |
| Sep 13, 2017 at 8:00 | review | First posts | |||
| Sep 13, 2017 at 8:02 | |||||
| Sep 13, 2017 at 7:56 | history | asked | Yoni Keren | CC BY-SA 3.0 |