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
    • General Computer Science

    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