New developments of the Z-EDM algorithm

In this paper we address some open questions on the recently proposed Zero-Error Density Maximization algorithm for MLP training. We propose a new version of the cost function that solves a training problem encountered in previous work and prove that the use of a nonparametric density estimator pres...

Full description

Saved in:
Bibliographic Details
Published inSixth International Conference on Intelligent Systems Design and Applications Vol. 1; pp. 1067 - 1072
Main Authors Silva, L.M., de Sa, J.M., Alexandre, L.A.
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.10.2006
Subjects
Cost function
Encoding
Entropy
Error correction
Hypercubes
Intelligent systems
Kernel
Multilayer perceptrons
Pattern recognition
Smoothing methods
Online AccessGet full text
ISBN0769525288
9780769525280
ISSN2164-7143
DOI10.1109/ISDA.2006.204

Cover

More Information
Summary:In this paper we address some open questions on the recently proposed Zero-Error Density Maximization algorithm for MLP training. We propose a new version of the cost function that solves a training problem encountered in previous work and prove that the use of a nonparametric density estimator preserves the optimal solution. Some experiments are reported comparing this cost function to the usual mean-square error and cross entropy cost functions
ISBN:0769525288
9780769525280
ISSN:2164-7143
DOI:10.1109/ISDA.2006.204