Control of Overfitting with Physics

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Kozyrev, Sergei V., Lopatin, Ilya A, Pechen, Alexander N
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929630176346112
author Kozyrev, Sergei V.
Lopatin, Ilya A
Pechen, Alexander N
author_facet Kozyrev, Sergei V.
Lopatin, Ilya A
Pechen, Alexander N
contents While there are many works on the applications of machine learning, not so many of them are trying to understand the theoretical justifications to explain their efficiency. In this work, overfitting control (or generalization property) in machine learning is explained using analogies from physics and biology. For stochastic gradient Langevin dynamics, we show that the Eyring formula of kinetic theory allows to control overfitting in the algorithmic stability approach - when wide minima of the risk function with low free energy correspond to low overfitting. For the generative adversarial network (GAN) model, we establish an analogy between GAN and the predator-prey model in biology. An application of this analogy allows us to explain the selection of wide likelihood maxima and overfitting reduction for GANs.
format Preprint
id arxiv_https___arxiv_org_abs_2412_10716
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Control of Overfitting with Physics
Kozyrev, Sergei V.
Lopatin, Ilya A
Pechen, Alexander N
Machine Learning
Artificial Intelligence
While there are many works on the applications of machine learning, not so many of them are trying to understand the theoretical justifications to explain their efficiency. In this work, overfitting control (or generalization property) in machine learning is explained using analogies from physics and biology. For stochastic gradient Langevin dynamics, we show that the Eyring formula of kinetic theory allows to control overfitting in the algorithmic stability approach - when wide minima of the risk function with low free energy correspond to low overfitting. For the generative adversarial network (GAN) model, we establish an analogy between GAN and the predator-prey model in biology. An application of this analogy allows us to explain the selection of wide likelihood maxima and overfitting reduction for GANs.
title Control of Overfitting with Physics
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2412.10716