Preprint
Aug 2026
When Single-Dataset Conclusions Fail: A 45-Task Study of Threshold Tuning and Resampling for Imbalanced Classification
Class-imbalance handling is routinely evaluated on a single benchmark dataset, and the resulting conclusions are reported as if they were properties of the method, and it is shown this practice is unsafe and, for a well-calibrated ensemble, imbalance handling is unnecessary.
Diyorbek Musaev
· 0 citations