Regularization of extreme learning machines with information of spatial relations of the projected data

Carregando...
Imagem de Miniatura

Data

Título da Revista

ISSN da Revista

Título de Volume

Editor

Universidade Federal de Minas Gerais

Descrição

Tipo

Artigo de evento

Título alternativo

Primeiro orientador

Membros da banca

Resumo

The following work presents a new approach to automatic selection of Tikhonov's regularization parameter, responsible for controlling the weight value of an ELM neural network. A strategy based on measurements obtained from data projection (Fisher-Score) is introduced. Seven datasets are tested and results are compared to those obtained when the regularization parameter is selected through cross-validation. The strategy shows satisfactory classification performance (in terms of p-value), while presenting significant training time reduction.

Abstract

Assunto

Redes neurais (Computação)

Palavras-chave

Neurons , Training , Cybernetics , Minimization , Testing , Biological neural networks , Aerospace electronics, Extreme Learning Machine , Training Time , Parameter Selection , Weight Decay , Regularization Parameter , Null Hypothesis , High-dimensional , Dimensional Space , Hidden Layer , Weight Matrix , Linear Discriminant Analysis , Weight Vector , Decision Boundary , Neurons In The Hidden Layer , Hidden Neurons , Affinity Matrix , Linearly Separable , Regularization Scheme , Class Mean , Universal Approximation , Norm Penalty , Fisher Score , Single Layer Feedforward Network , Empirical Risk Minimization

Citação

Curso

Endereço externo

https://ieeexplore.ieee.org/document/8820610

Avaliação

Revisão

Suplementado Por

Referenciado Por