Power Transform Revisited: Numerically Stable, and Federated

Xuefeng Xu, Graham Cormode
Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, PMLR 300:2800-2808, 2026.

Abstract

Power transforms are popular parametric methods for making data more Gaussian-like, and are widely used as preprocessing steps in statistical analysis and machine learning. However, we find that direct implementations of power transforms suffer from severe numerical instabilities, which can lead to incorrect results or even crashes. In this paper, we provide a comprehensive analysis of the sources of these instabilities and propose effective remedies. We further extend power transforms to the federated learning setting, addressing both numerical and distributional challenges that arise in this context. Experiments on real-world datasets demonstrate that our methods are both effective and robust, substantially improving stability compared to existing approaches.

Cite this Paper

BibTeX
@InProceedings{pmlr-v300-xu26d, title = { Power Transform Revisited: Numerically Stable, and Federated }, author = {Xu, Xuefeng and Cormode, Graham}, booktitle = {Proceedings of The 29th International Conference on Artificial Intelligence and Statistics}, pages = {2800--2808}, year = {2026}, editor = {Khan, Emtiyaz and Li, Yingzhen and Solin, Arno and Ramdas, Aaditya}, volume = {300}, series = {Proceedings of Machine Learning Research}, month = {02--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v300/main/assets/xu26d/xu26d.pdf}, url = {https://proceedings.mlr.press/v300/xu26d.html}, abstract = { Power transforms are popular parametric methods for making data more Gaussian-like, and are widely used as preprocessing steps in statistical analysis and machine learning. However, we find that direct implementations of power transforms suffer from severe numerical instabilities, which can lead to incorrect results or even crashes. In this paper, we provide a comprehensive analysis of the sources of these instabilities and propose effective remedies. We further extend power transforms to the federated learning setting, addressing both numerical and distributional challenges that arise in this context. Experiments on real-world datasets demonstrate that our methods are both effective and robust, substantially improving stability compared to existing approaches. } }
Endnote
%0 Conference Paper %T Power Transform Revisited: Numerically Stable, and Federated %A Xuefeng Xu %A Graham Cormode %B Proceedings of The 29th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2026 %E Emtiyaz Khan %E Yingzhen Li %E Arno Solin %E Aaditya Ramdas %F pmlr-v300-xu26d %I PMLR %P 2800--2808 %U https://proceedings.mlr.press/v300/xu26d.html %V 300 %X Power transforms are popular parametric methods for making data more Gaussian-like, and are widely used as preprocessing steps in statistical analysis and machine learning. However, we find that direct implementations of power transforms suffer from severe numerical instabilities, which can lead to incorrect results or even crashes. In this paper, we provide a comprehensive analysis of the sources of these instabilities and propose effective remedies. We further extend power transforms to the federated learning setting, addressing both numerical and distributional challenges that arise in this context. Experiments on real-world datasets demonstrate that our methods are both effective and robust, substantially improving stability compared to existing approaches.
APA
Xu, X. & Cormode, G.. (2026). Power Transform Revisited: Numerically Stable, and Federated . Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 300:2800-2808 Available from https://proceedings.mlr.press/v300/xu26d.html.

Related Material