Timeline for Using summary statistics from full dataset for feature selection
Current License: CC BY-SA 4.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 23, 2018 at 8:15 | vote | accept | Robert Patchouli | ||
| Nov 22, 2018 at 17:26 | comment | added | Jespar | When using CV the idea is to randomise on each iteration the training set and the valuation set, while keeping their sizes the same 60/30 (and 10 for testing). Or whatever other ratio you prefer. I wouldn't touch the prediction dataset. | |
| Nov 22, 2018 at 17:22 | comment | added | Robert Patchouli | If I use CV, I still have a "prediction" dataset, can I use the information from that dataset for preprocessing decisions? | |
| Nov 22, 2018 at 17:18 | history | answered | Jespar | CC BY-SA 4.0 |