Fast learning fully complex-valued classifiers for real-valued classification problems

R. Savitha, S. Suresh, N. Sundararajan, H. J. Kim

Research output: Chapter in Book/Report/Conference proceedingConference contribution

28 Citations (Scopus)

Abstract

In this paper, we present two fast learning neural network classifiers with a single hidden layer: the 'Phase Encoded Complex-valued Extreme Learning Machine (PE-CELM)' and the 'Bilinear Branch-cut Complex-valued Extreme Learning Machine (BB-CELM)'. The proposed classifiers use the phase encoded transformation and the bilinear transformation with a branch-cut at 2π as the activation functions in the input layer to map the real-valued features to the complex domain. The neurons in the hidden layer employ the fully complex-valued activation function of the type of a hyperbolic secant function. The parameters of the hidden layer are chosen randomly and the output weights are estimated as the minimum norm least square solution to a set of linear equations. The classification ability of these classifiers are evaluated using a set of benchmark data sets from the UCI machine learning repository. Results highlight the superior classification ability of these classifiers with least computational effort.

Original languageEnglish
Title of host publicationAdvances in Neural Networks - 8th International Symposium on Neural Networks, ISNN 2011
Pages602-609
Number of pages8
EditionPART 1
DOIs
Publication statusPublished - 2011
Event8th International Symposium on Neural Networks, ISNN 2011 - Guilin, China
Duration: 2011 May 292011 Jun 1

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 1
Volume6675 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other8th International Symposium on Neural Networks, ISNN 2011
Country/TerritoryChina
CityGuilin
Period11/5/2911/6/1

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Fast learning fully complex-valued classifiers for real-valued classification problems'. Together they form a unique fingerprint.

Cite this