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

arXiv: https://arxiv.org/abs/2501.15790 · HTML: https://mlsanity.com/paper/2501.15790/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 4 of its 51 references that are in the map.
- First posted on arXiv 27 Jan 2025.

## Builds on (most cited first, 4 of 4)

- [How Faithful is your Synthetic Data? Sample-level Metrics for Evaluating and Auditing Generative Models](https://mlsanity.com/paper/2102.08921/) (2021)
- [Imbalanced-learn: A Python Toolbox to Tackle the Curse of Imbalanced Datasets in Machine Learning](https://mlsanity.com/paper/1609.06570/) (2016)
- [To SMOTE, or not to SMOTE?](https://mlsanity.com/paper/2201.08528/) (2022)
- [Foundations of data imbalance and solutions for a data democracy](https://mlsanity.com/paper/2108.00071/) (2021)

## Built on by (most cited first, 1 of 1)

- [Kolmogorov Arnold Networks (KANs) for Imbalanced Data -- An Empirical Perspective](https://mlsanity.com/paper/2507.14121/) (2025)

## Abstract (arXiv)

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, established for each instance by construction rather than assumed. We introduce Certified Interpolation Safe Oversampling (CISO), a three-phase interpolation framework built for this objective. A safety-guided distribution selects where synthesis occurs, a locality-and clearance-weighted distribution selects with whom each anchor interpolates, and a q-Gaussian placement density determines how far along the resulting segment each instance is placed. The framework provides three guarantees. Each synthetic instance seeded by a sufficiently safe anchor carries a certified distance from the majority class; a signed temperature parameter provably and monotonically shifts synthesis between boundary-seeking and interior-seeking regimes; and selection weights are strictly positive by construction, so no degenerate case arises. Certification is obtained alongside competitive predictive performance rather than in place of it. Under a protocol preregistered before evaluation, across 45 datasets, four classifiers, and eleven competing methods, CISO is statistically equivalent to SMOTE on precision-recall AUC, ranks second of eleven under gradient boosting, and completes every one of 11,460 fold-level evaluations without failure. A parameter sweep further reveals a continuous fidelity-safety trade-off that competing methods occupy only as isolated points.