Multiclass threshold-based classification

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Marchetti, Francesco, Legnaro, Edoardo, Guastavino, Sabrina
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914175393988608
author Marchetti, Francesco
Legnaro, Edoardo
Guastavino, Sabrina
author_facet Marchetti, Francesco
Legnaro, Edoardo
Guastavino, Sabrina
contents In this paper, we introduce a threshold-based framework for multiclass classification that generalizes the standard argmax rule. This is done by replacing the probabilistic interpretation of softmax outputs with a geometric one on the multidimensional simplex, where the classification depends on a multidimensional threshold. This change of perspective enables for any trained classification network an a posteriori optimization of the classification score by means of threshold tuning, as usually carried out in the binary setting. This allows a further refinement of the prediction capability of any network. Moreover, this multidimensional threshold-based setting makes it possible to define score-oriented losses, which are based on the interpretation of the threshold as a random variable. Our experiments show that the multidimensional threshold tuning yields consistent performance improvements across various networks and datasets, and that the proposed multiclass score-oriented losses are competitive with standard loss functions, resembling the advantages observed in the binary case.
format Preprint
id arxiv_https___arxiv_org_abs_2505_11276
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Multiclass threshold-based classification
Marchetti, Francesco
Legnaro, Edoardo
Guastavino, Sabrina
Machine Learning
In this paper, we introduce a threshold-based framework for multiclass classification that generalizes the standard argmax rule. This is done by replacing the probabilistic interpretation of softmax outputs with a geometric one on the multidimensional simplex, where the classification depends on a multidimensional threshold. This change of perspective enables for any trained classification network an a posteriori optimization of the classification score by means of threshold tuning, as usually carried out in the binary setting. This allows a further refinement of the prediction capability of any network. Moreover, this multidimensional threshold-based setting makes it possible to define score-oriented losses, which are based on the interpretation of the threshold as a random variable. Our experiments show that the multidimensional threshold tuning yields consistent performance improvements across various networks and datasets, and that the proposed multiclass score-oriented losses are competitive with standard loss functions, resembling the advantages observed in the binary case.
title Multiclass threshold-based classification
topic Machine Learning
url https://arxiv.org/abs/2505.11276