Title :
Application of the information-Theoretical approach to multilayer perceptron training
Author :
Morozov, O.A. ; Ovchinnikov, P.E.
Author_Institution :
Dept. of Phys., Nizhny Novgorod State Univ., Nizhny Novgorod, Russia
Abstract :
The effective method for training of multilayer perceptron network (MLP), consisting in optimization of information criterion on the basis of Renyi´s entropy, is proposed. The results of computer modeling are presented. Training speed increase by the proposed scheme in comparison with the traditional approach is demonstrated experimentally.
Keywords :
entropy; learning (artificial intelligence); multilayer perceptrons; optimisation; Renyi entropy; information theory; multilayer perceptron network; optimization; training; Probability; Training; Renyi´s entropy; artificial neural network; training;
Conference_Titel :
Communication Software and Networks (ICCSN), 2011 IEEE 3rd International Conference on
Conference_Location :
Xi´an
Print_ISBN :
978-1-61284-485-5
DOI :
10.1109/ICCSN.2011.6014851
