Is it possible to use binary cross etropy loss with non binary ground truth data, i.e. not [0,1] values, but [0,0.1,0.5,1.0] values.
- Does it make sense?
- Does it have any side effects?
Asked
Active
Viewed 187 times
1
mrgloom
- 1,687
- 4
- 25
- 33
-
Could you explain (a) what "binary cross entropy loss" would actually be with such data; (b) exactly what your data are measuring; and (c) what you mean by "use"? – whuber Dec 03 '18 at 18:08
-
This topics answer my question https://stats.stackexchange.com/a/296277 https://stackoverflow.com/a/52443301 – mrgloom Dec 04 '18 at 00:24