Skip to content
arXiv stat.ML · Papers

Certified Interpolation Oversampling: Per-Instance Safety Guarantees for Imbalanced Learning

arXiv:2501.15790v2 Announce Type: replace-cross Abstract: Synthetic minority oversampling is typically designed and evaluated against a predictive objective, generating samples that improve downstream classification. This paper pursues a second objective by generating samples that carry a stated safety property, establ