Title Logistic regression using covariates obtained by product-unit neural network models
Authors Hervás-Martínez C., MARTÍNEZ ESTUDILLO, FRANCISCO JOSÉ, MARTÍNEZ ESTUDILLO, FRANCISCO JOSÉ
External publication No
Means Pattern Recogn.
Scope Article
Nature Científica
JCR Quartile 1
SJR Quartile 1
JCR Impact 2.01900
SJR Impact 1.27500
Area International
Web https://www.scopus.com/inward/record.uri?eid=2-s2.0-33749267856&doi=10.1016%2fj.patcog.2006.06.003&partnerID=40&md5=3ea48011278191c5dba0f2a5291fa27f
Publication date 01/01/2007
ISI 000241837300005
Scopus Id 2-s2.0-33749267856
DOI 10.1016/j.patcog.2006.06.003
Abstract We propose a logistic regression method based on the hybridation of a linear model and product-unit neural network models for binary classification. In a first step we use an evolutionary algorithm to determine the basic structure of the product-unit model and afterwards we apply logistic regression in the new space of the derived features. This hybrid model has been applied to seven benchmark data sets and a new microbiological problem. The hybrid model outperforms the linear part and the nonlinear part obtaining a good compromise between them and they perform well compared to several other learning classification techniques. We obtain a binary classifier with very promising results in terms of classification accuracy and the complexity of the classifier. © 2006 Pattern Recognition Society.
Keywords Classification (of information); Computational complexity; Evolutionary algorithms; Mathematical models; Neural networks; Problem solving; Benchmark data sets; Logistic regression; Product-unit neural
Universidad Loyola members