GMDH Method with Genetic Selection Algorithm and Cloning

0421992 - ÚI 2014 RIV CZ eng J - Journal Article
Jiřina, Marcel - Jiřina jr., M.
GMDH Method with Genetic Selection Algorithm and Cloning.
Neural Network World. Roč. 23, č. 5 (2013), s. 451-464. ISSN 1210-0552
Institutional support: RVO:67985807
Keywords : multivariate data * GMDH * linear regression * Gauss-Markov conditions * cloning * genetic selection * classification
Subject RIV: BB - Applied Statistics, Operational Research
Impact factor: 0.412, year: 2013

The GMDH MIA algorithm uses linear regression for adaptation. We show that Gauss-Markov conditions are not met here and thus estimations of network parameters are biased. To eliminate this we propose to use cloning of neuron parameters in the GMDH network with genetic selection and cloning (GMC GMDH) that can outperform other powerful methods. It is demonstrated on tasks from the Machine Learning Repository.
Permanent Link: http://hdl.handle.net/11104/0228224