[2508.10053] xRFM: Accurate, scalable, and interpretable feature learning models for tabular data

Computer Science > Machine Learning arXiv:2508.10053 (cs) [Submitted on 12 Aug 2025 (v1), last revised 5 Apr 2026 (this version, v3)] Title:xRFM: Accurate, scalable, and interpretable feature learning models for tabular data Authors:Daniel Beaglehole, David Holzmüller, Adityanarayanan Radhakrishnan, Mikhail Belkin View a PDF of the paper titled xRFM: Accurate, scalable, and interpretable feature learning models for tabular data, by Daniel Beaglehole and 3 other authors View PDF HTML (experimental) Abstract:Inference from tabular data, collections of continuous and categorical variables organized into matrices, is a foundation for modern technology and science. Yet, in contrast to the explosive changes in the rest of AI, the best practice for these predictive tasks has been relatively unchanged and is still primarily based on variations of Gradient Boosted Decision Trees (GBDTs). Very recently, there has been renewed interest in developing state-of-the-art methods for tabular data based on recent developments in neural networks and feature learning methods. In this work, we introduce xRFM, an algorithm that combines feature learning kernel machines with a tree structure to both adapt to the local structure of the data and scale to essentially unlimited amounts of training data. We show that compared to $31$ other methods, including recently introduced tabular foundation models (TabPFNv2) and GBDTs, xRFM achieves best performance across $100$ regression datasets and is compe...