Initialization of Supervised Training for Parametric Estimation

Neural Processing Letters

Volumn 9, Issue 1, 1999, Pages 53-61

  Costa, P ^a   Larzabal, P ^a  

^a UNIVERSITÉ PARIS SACLAY   (France)

Author keywords

Estimation; Global convergence; Initialization; Multilayer perceptron

Indexed keywords

APPROXIMATION THEORY; BACKPROPAGATION; CONVERGENCE OF NUMERICAL METHODS; LEARNING ALGORITHMS; LEARNING SYSTEMS; PARAMETER ESTIMATION; RANDOM PROCESSES;

SUPERVISED TRAINING; TAYLOR DEVELOPMENT; TRAINING ALGORITHMS;

MULTILAYER NEURAL NETWORKS;

EID: 0033075788     PISSN: 13704621     EISSN: None     Source Type: Journal    
DOI: 10.1023/A:1018671912219     Document Type: Article

References (5)

1. P. Costa , P. Larzabal and H. Clergeot, "On Initialization of Neural Network", Proceedings of ICNN 95, Perth, Australia, 29 Nov.-1 Dec., 1995.
2. D. Nguyen and B. Widrow, "Improving the Learning Speed of 2-Layer Neural Networks by choosing Initial Values of the Adaptative Weights", Proc. Intl. Joint Conf. on Neural Networks, San Diego, Calif., June 1990.
3. S. Osowski, "New approach to selection of initial values of weights in neural function approximation", Electronics Letters, Vol. 29, No. 3, pp. 313-315, Feb. 1993.
4. T. Denoeux and R. Lengellé, "Initializing back propagation networks with prototypes", Neural Networks, Vol. 6, pp. 351-363, 1993.
5. G. Cybenko, "Approximation by superpositions of a sigmoidal function", Mathematics of Control, Signal and Systems, Vol. 2, pp. 303-314, 1989.