Ideal step size estimation for the multinomial logistic regression

At the core of deep learning optimization problems reside algorithms such as the Stochastic Gradient Descent (SGD), which employs a subset of the data per iteration to estimate the gradient in order to minimize a cost function. Adaptive algorithms, based on SGD, are well known for being effective in...

Descripción completa

Detalles Bibliográficos
Autor: Ramirez Orihuela, Gabriel
Tipo de recurso: tesis de maestría
Fecha de publicación:2024
País:Perú
Institución:Pontificia Universidad Católica del Perú
Repositorio:PUCP-Tesis
Idioma:inglés
OAI Identifier:oai:tesis.pucp.edu.pe:20.500.12404/29791
Acceso en línea:http://hdl.handle.net/20.500.12404/29791
Access Level:acceso abierto
Palabra clave:Aprendizaje automático (Inteligencia artificial)
Aprendizaje profundo (Aprendizaje automático)
Optimización matemática
Análisis de regresión
https://purl.org/pe-repo/ocde/ford#2.00.00
Descripción
Sumario:At the core of deep learning optimization problems reside algorithms such as the Stochastic Gradient Descent (SGD), which employs a subset of the data per iteration to estimate the gradient in order to minimize a cost function. Adaptive algorithms, based on SGD, are well known for being effective in using gradient information from past iterations, generating momentum or memory that enables a more accurate prediction of the true gradient slope in future iterations, thus accelerating convergence. Nevertheless, these algorithms still need an initial (scalar) learning rate (LR) as well as a LR scheduler. In this work we propose a new SGD algorithm that estimates the initial (scalar) LR via an adaptation of the ideal Cauchy step size for the multinomial logistic regression; furthermore, the LR is recursively updated up to a given number of epochs, after which a decaying LR scheduler is used. The proposed method is assessed for several well-known multiclass classification architectures and favorably compares against other well-tuned (scalar and spatially) adaptive alternatives, including the Adam algorithm.