Catalogo Articoli (Spogli Riviste)
OPAC HELP
Titolo: Time series analysis using normalized PGRBF network with regression weights
Autore: Rojas, I; Pomares, H; Bernier, JL; Ortega, J; Pino, B; Pelayo, FJ; Prieto, A;
 Indirizzi:
 Univ Granada, Dept Architecture & Comp Technol, E18071 Granada, Spain Univ Granada Granada Spain E18071 Comp Technol, E18071 Granada, Spain
 Titolo Testata:
 NEUROCOMPUTING
,
volume: 42,
anno: 2002,
pagine: 267  285
 SICI:
 09252312(200201)42:<267:TSAUNP>2.0.ZU;23
 Fonte:
 ISI
 Lingua:
 ENG
 Soggetto:
 PREDICTION;
 Keywords:
 RBF neural networks; sequential learning; pruning strategy; network growing; time series prediction;
 Tipo documento:
 Article
 Natura:
 Periodico
 Settore Disciplinare:
 Engineering, Computing & Technology
 Citazioni:
 20
 Recensione:
 Indirizzi per estratti:
 Indirizzo: Rojas, I Univ Granada, Dept Architecture & Comp Technol, Campus Univ Fuentenueva, E18071 Granada, Spain Univ Granada Campus Univ Fuentenueva Granada Spain E18071 Spain



 Citazione:
 I. Rojas et al., "Time series analysis using normalized PGRBF network with regression weights", NEUROCOMPUT, 42, 2002, pp. 267285
Abstract
This paper proposes a framework for constructing and training a radial basis function (RBF) neural network. For this purpose, a sequential learning algorithm is presented to adapt the structure of the network, in which it ispossible to create a new hidden unit and also to detect and remove inactive units. The structure of the Gaussian functions is modified using a pseudoGaussian function (PG) in which two scaling parameters a are introduced, which eliminates the symmetry restriction and provides the neurons in the hidden layer with greater flexibility with respect to function approximation. Other important characteristics of the proposed neural system are that theactivation of the hidden neurons is normalized which, as described in the bibliography, provides a better performance than nonnormalization and instead of using a single parameter for the output weights, these are functions of the input variables which leads to a significant reduction in the numberof hidden units compared to the classical RBF network. Finally, we examinethe result of applying the proposed algorithm to time series prediction. (C) 2002 Elsevier Science B.V. All rights reserved.
ASDD Area Sistemi Dipartimentali e Documentali, Università di Bologna, Catalogo delle riviste ed altri periodici
Documento generato il 07/07/20 alle ore 18:56:33