Optimal downsampling for Imbalanced Classification with Generalized Linear Models

Yan Chen, Jose Blanchet, Krzysztof Dembczynski, Laura Fee Nern, Aaron Eliasib Flores
Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, PMLR 258:1306-1314, 2025.

Abstract

Downsampling or under-sampling is a technique that is utilized in the context of large and highly imbalanced classification models. We study optimal downsampling for imbalanced classification using generalized linear models (GLMs). We propose a pseudo maximum likelihood estimator and study its asymptotic normality in the context of increasingly imbalanced populations relative to an increasingly large sample size. We provide theoretical guarantees for the introduced estimator. Additionally, we compute the optimal downsampling rate using a criterion that balances statistical accuracy and computational efficiency. Our numerical experiments, conducted on both synthetic and empirical data, further validate our theoretical results, and demonstrate that the introduced estimator outperforms commonly available alternatives.

Cite this Paper

BibTeX
@InProceedings{pmlr-v258-chen25c, title = {Optimal downsampling for Imbalanced Classification with Generalized Linear Models}, author = {Chen, Yan and Blanchet, Jose and Dembczynski, Krzysztof and Nern, Laura Fee and Flores, Aaron Eliasib}, booktitle = {Proceedings of The 28th International Conference on Artificial Intelligence and Statistics}, pages = {1306--1314}, year = {2025}, editor = {Li, Yingzhen and Mandt, Stephan and Agrawal, Shipra and Khan, Emtiyaz}, volume = {258}, series = {Proceedings of Machine Learning Research}, month = {03--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v258/main/assets/chen25c/chen25c.pdf}, url = {https://proceedings.mlr.press/v258/chen25c.html}, abstract = {Downsampling or under-sampling is a technique that is utilized in the context of large and highly imbalanced classification models. We study optimal downsampling for imbalanced classification using generalized linear models (GLMs). We propose a pseudo maximum likelihood estimator and study its asymptotic normality in the context of increasingly imbalanced populations relative to an increasingly large sample size. We provide theoretical guarantees for the introduced estimator. Additionally, we compute the optimal downsampling rate using a criterion that balances statistical accuracy and computational efficiency. Our numerical experiments, conducted on both synthetic and empirical data, further validate our theoretical results, and demonstrate that the introduced estimator outperforms commonly available alternatives.} }
Endnote
%0 Conference Paper %T Optimal downsampling for Imbalanced Classification with Generalized Linear Models %A Yan Chen %A Jose Blanchet %A Krzysztof Dembczynski %A Laura Fee Nern %A Aaron Eliasib Flores %B Proceedings of The 28th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2025 %E Yingzhen Li %E Stephan Mandt %E Shipra Agrawal %E Emtiyaz Khan %F pmlr-v258-chen25c %I PMLR %P 1306--1314 %U https://proceedings.mlr.press/v258/chen25c.html %V 258 %X Downsampling or under-sampling is a technique that is utilized in the context of large and highly imbalanced classification models. We study optimal downsampling for imbalanced classification using generalized linear models (GLMs). We propose a pseudo maximum likelihood estimator and study its asymptotic normality in the context of increasingly imbalanced populations relative to an increasingly large sample size. We provide theoretical guarantees for the introduced estimator. Additionally, we compute the optimal downsampling rate using a criterion that balances statistical accuracy and computational efficiency. Our numerical experiments, conducted on both synthetic and empirical data, further validate our theoretical results, and demonstrate that the introduced estimator outperforms commonly available alternatives.
APA
Chen, Y., Blanchet, J., Dembczynski, K., Nern, L.F. & Flores, A.E.. (2025). Optimal downsampling for Imbalanced Classification with Generalized Linear Models. Proceedings of The 28th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 258:1306-1314 Available from https://proceedings.mlr.press/v258/chen25c.html.

Related Material