Paper Overview
Field: Machine Learning Authors: Shady E. Ahmed, Panos Stinis Published: 2026-07-29 arXiv: 2607.27196
Summary
This paper introduces a novel regression method inspired by the mechanism fruitflies use to sense their environment: performing regression via classification. The authors build a general framework for learning nonlinear input-output relationships by replacing complex global surrogate models with a finite library of representative local patterns.
Since scientific data often occupy limited and recurring regions of the input space, predictions are generated by measuring the similarity between a query and stored patterns, then combining their associated responses through weighted reconstruction.
The method is applied to nonlinear dynamical systems, data-driven regression, and physics-informed learning, using suitable embeddings and similarity measures. For dynamical systems, the proposed offline-online workflow extracts patterns offline from data or governing equations, while online prediction only requires similarity evaluation and response aggregation. This structure helps reduce computational and memory requirements while providing explicit control over the trade-off between accuracy, storage, and inference cost.
Original Abstract (excerpt)
> We present a novel approach to regression tasks using classification which is motivated by the mechanism used by fruitflies to sense their environment. Specifically, we formulate a general framework for learning nonlinear input-output relationships by replacing complex global surrogate models with a finite library of representative local patterns. Since scientific data often occupy limited and recurring regions of the input space, we generate predictions by measuring similarities between a query and stored patterns, then combining their associated responses through weighted reconstruction. We apply this approach to nonlinear dynamical systems, data-driven regression, and physics-informed learning using suitable embeddings and similarity measures. For dynamical systems, our offline-online w...
---
*Auto-collected on 2026-07-31*