Timeline for Why not just dump the neural networks and deep learning?
Current License: CC BY-SA 3.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 25, 2018 at 12:47 | history | made wiki | Post Made Community Wiki by whuber♦ | ||
| Aug 13, 2017 at 8:42 | comment | added | Miguel | @RajeshDachiraju The same could be said on penalty coefficients in exterior point optimization algorithms, or the step size in Runge-Kutta methods. The word "inconsistent" has a precise meaning in science that does not apply here. | |
| Aug 13, 2017 at 0:52 | comment | added | Rajesh D | Problem with being inconsistent is that, one cannot ask simple questions like, When should one stop training and give up? Also lot of rumours like, 'Dropot', 'weight decay', 'ReLu' and various activations, batch normalization, max pooling, softmax, early stopping, various learning rate schedules and all permutations and combinations of these make the designer always in doubt whether to give up or not at some point. | |
| Aug 11, 2017 at 16:27 | review | First posts | |||
| Aug 11, 2017 at 17:04 | |||||
| Aug 11, 2017 at 16:22 | history | answered | Miguel | CC BY-SA 3.0 |