%X This paper investigates the performance of conjugate gradient algorithms with sliding-window approach for training multilayer perceptron (MLP). Online learning is implemented when the system under investigation is time varying or when it is not convenient to obtain a full history of offline data about the system variables. Sliding window framework is proposed to combine the robustness of offline learning with the ability of online learning to track time varying elements of the process under investigation. A sliding window based second order conjugate gradient algorithms SWCG is presented. The performance of SWCG is compared with a sliding window based first order back propagation SWBP. © 2011 IEEE.
%K Conjugate gradient; Conjugate gradient algorithms; First order; Multi layer perceptron; Nonlinear conjugate gradient; Off-line learning; Offline data; Online learning; Second orders; Sliding window-based; Sliding-window; System variables; Time varying, Backpropagation algorithms; Conjugate gradient method; Neural networks; Signal processing; Time varying networks, E-learning
%L scholars2081
%J Proceedings - 2011 IEEE 7th International Colloquium on Signal Processing and Its Applications, CSPA 2011
%O cited By 8; Conference of 2011 IEEE 7th International Colloquium on Signal Processing and Its Applications, CSPA 2011 ; Conference Date: 4 March 2011 Through 6 March 2011; Conference Code:84971
%R 10.1109/CSPA.2011.5759854
%D 2011
%A H. Izzeldin
%A V.S. Asirvadam
%A N. Saad
%T Online sliding-window based for training MLP networks using advanced conjugate gradient
%C Penang
%P 112-116