Timeline for Fitting model on whole dataset, more or less epochs ? (w.r.t validation accuracy) [duplicate]
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 25, 2019 at 4:14 | history | closed |
Jan Kukacka kjetil b halvorsen♦ jbowman Ferdi StatsStudent |
Duplicate of Is epoch optimization in CV with constant mini-batch size even possible? | |
| Jan 8, 2019 at 12:55 | comment | added | DeltaIV | @JanKukacka it's not correct to say that you cannot use this method. It's more correct to say that in some cases it works, and in some others it doesn't, and you should be trying to keep the same number of parameter updates, rather than the same number of epochs. See deeplearningbook.org/contents/regularization.html, paragraph 7.8, and arxiv.org/pdf/1206.5533v2.pdf | |
| Jan 8, 2019 at 11:40 | review | Close votes | |||
| Jan 25, 2019 at 4:14 | |||||
| Jan 8, 2019 at 11:26 | comment | added | Jan Kukacka | As discussed in the above linked thread, you cannot use this method as the number of actual weight updates per epoch (gradient descent steps) changes if you change the size of the training set. Also, the benefit of having a few more samples might not outweight the risk of overfitting, but this depends on the dataset size and task complexity... | |
| Jan 7, 2019 at 16:02 | answer | added | DeltaIV | timeline score: 5 | |
| Jan 7, 2019 at 12:59 | history | asked | Isbister | CC BY-SA 4.0 |