Title :
A Hessian matrix approach for training nonlinear networks
Author :
Yu, Changhua ; Manry, M.T.
Author_Institution :
Dept. of Electr. Eng., Texas Univ., Arlington, TX, USA
fDate :
31 Aug.-4 Sept. 2004
Abstract :
In the original output weight optimization-hidden weight optimization (OWO-HWO) algorithm for training multilayer perceptions, only first order information is used to construct the desired net function. This gradient-like strategy inevitably reduces efficiency. In this paper, an efficient Hessian matrix inversion method is proposed for the hidden weights optimization. Numerical results validate the improvement of this algorithm.
Keywords :
Hessian matrices; gradient methods; learning (artificial intelligence); matrix inversion; multilayer perceptrons; optimisation; Hessian matrix approach; gradient-like strategy; multilayer perception; nonlinear network training; output weight optimization-hidden weight optimization algorithm; Convergence; Delay; Joining processes; Multilayer perceptrons; Nonlinear equations; Optimization methods; Remote sensing; Signal processing; Signal processing algorithms; Training data;
Conference_Titel :
Signal Processing, 2004. Proceedings. ICSP '04. 2004 7th International Conference on
Print_ISBN :
0-7803-8406-7
DOI :
10.1109/ICOSP.2004.1441615
