I have 36168 data with imbalanced target. 88,3% is 0 (31970 data) and 11,7% is 1 (4198 data). I want to apply oversampling using SMOTE. Is it ideal to make it the same amount of data so the 0 & 1 target contains 31970 data? Because i think in the industry there will be no data with perfect balanced amount of target. Or should i using sampling_strategy parameter (make it 0.8 / 0.7 ratio)?
Asked
Active
Viewed 184 times
0
-
[There are a number of common misconceptions about class imbalance, chief among them being that artificial balancing is needed.](https://stats.meta.stackexchange.com/q/6349/247274) Thus, why balance at all? – Dave Oct 12 '22 at 17:19