$f$-Divergence Based Classification: Beyond the Use of Cross-Entropy

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Novello, Nicola, Tonello, Andrea M.
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916248469635072
author Novello, Nicola
Tonello, Andrea M.
author_facet Novello, Nicola
Tonello, Andrea M.
contents In deep learning, classification tasks are formalized as optimization problems often solved via the minimization of the cross-entropy. However, recent advancements in the design of objective functions allow the usage of the $f$-divergence to generalize the formulation of the optimization problem for classification. We adopt a Bayesian perspective and formulate the classification task as a maximum a posteriori probability problem. We propose a class of objective functions based on the variational representation of the $f$-divergence. Furthermore, driven by the challenge of improving the state-of-the-art approach, we propose a bottom-up method that leads us to the formulation of an objective function corresponding to a novel $f$-divergence referred to as shifted log (SL). We theoretically analyze the objective functions proposed and numerically test them in three application scenarios: toy examples, image datasets, and signal detection/decoding problems. The analyzed scenarios demonstrate the effectiveness of the proposed approach and that the SL divergence achieves the highest classification accuracy in almost all the considered cases.
format Preprint
id arxiv_https___arxiv_org_abs_2401_01268
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle $f$-Divergence Based Classification: Beyond the Use of Cross-Entropy
Novello, Nicola
Tonello, Andrea M.
Machine Learning
Signal Processing
In deep learning, classification tasks are formalized as optimization problems often solved via the minimization of the cross-entropy. However, recent advancements in the design of objective functions allow the usage of the $f$-divergence to generalize the formulation of the optimization problem for classification. We adopt a Bayesian perspective and formulate the classification task as a maximum a posteriori probability problem. We propose a class of objective functions based on the variational representation of the $f$-divergence. Furthermore, driven by the challenge of improving the state-of-the-art approach, we propose a bottom-up method that leads us to the formulation of an objective function corresponding to a novel $f$-divergence referred to as shifted log (SL). We theoretically analyze the objective functions proposed and numerically test them in three application scenarios: toy examples, image datasets, and signal detection/decoding problems. The analyzed scenarios demonstrate the effectiveness of the proposed approach and that the SL divergence achieves the highest classification accuracy in almost all the considered cases.
title $f$-Divergence Based Classification: Beyond the Use of Cross-Entropy
topic Machine Learning
Signal Processing
url https://arxiv.org/abs/2401.01268