Degree-Optimized Cumulative Polynomial Kolmogorov-Arnold Networks
Mathew Vanherreweghe, Lirandë Pira, Patrick Rebentrost · 2025
Read Degree-Optimized Cumulative Polynomial Kolmogorov-Arnold Networks by Mathew Vanherreweghe, Lirandë Pira, Patrick Rebentrost (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-15228.txt.
Description
Mathew Vanherreweghe, Lirandë Pira, and Patrick Rebentrost tune a polynomial neural network by turning its design choices into a puzzle for quantum-style optimizers, in a 2025 paper. Their networks build each neuron from Chebyshev polynomials, whose degrees must be chosen per neuron. Brute force over all choices explodes. Recasting degree choice as a binary optimization problem collapses it to one solve per layer.
The design shines on small datasets, shrugs off input scaling, and regularizes itself through its polynomial bones. Theory ties its behavior to traits of financial time series. Across several domains it matched or beat traditional architectures where data were scarce and stability mattered. Code and overhead tricks are published.
Catalog record
| CREATOR | Mathew Vanherreweghe, Lirandë Pira, Patrick Rebentrost |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-15228