Downsizing Machine Learning Models by Optimization through Ising Models

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autor principal: Okamoto, Yasuharu
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916662778789888
author Okamoto, Yasuharu
author_facet Okamoto, Yasuharu
contents This paper introduces a technique to enhance the efficiency of quadratic machine learning models, particularly Field-Aware Factorization Machines (FFMs) handling binary data. Our approach strategically reduces model size through optimized feature selection based on the Ising model, maintaining comparable accuracy to the original model. By exploiting the adjustability of FFM's cross-term weights during a novel two-stage training process, we demonstrate a significant improvement in overall computational efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2503_20123
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Downsizing Machine Learning Models by Optimization through Ising Models
Okamoto, Yasuharu
Materials Science
This paper introduces a technique to enhance the efficiency of quadratic machine learning models, particularly Field-Aware Factorization Machines (FFMs) handling binary data. Our approach strategically reduces model size through optimized feature selection based on the Ising model, maintaining comparable accuracy to the original model. By exploiting the adjustability of FFM's cross-term weights during a novel two-stage training process, we demonstrate a significant improvement in overall computational efficiency.
title Downsizing Machine Learning Models by Optimization through Ising Models
topic Materials Science
url https://arxiv.org/abs/2503.20123