TY - GEN
T1 - Fast learning fully complex-valued classifiers for real-valued classification problems
AU - Savitha, R.
AU - Suresh, S.
AU - Sundararajan, N.
AU - Kim, H. J.
PY - 2011
Y1 - 2011
N2 - 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.
AB - 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.
UR - http://www.scopus.com/inward/record.url?scp=79957799279&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-21105-8_70
DO - 10.1007/978-3-642-21105-8_70
M3 - Conference contribution
AN - SCOPUS:79957799279
SN - 9783642211041
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 602
EP - 609
BT - Advances in Neural Networks - 8th International Symposium on Neural Networks, ISNN 2011
T2 - 8th International Symposium on Neural Networks, ISNN 2011
Y2 - 29 May 2011 through 1 June 2011
ER -