An algorithm for fast convergence in training neural networks

An algorithm for fast convergence in training neural networks (2001)

Cached

Download Links

by Bogdan M. Wilamowski , Okyay Kaynak

Venue:	Proceedings of the International Joint Conference on Neural Networks, 2:1778–1782
Citations:	1 - 0 self

BibTeX

@INPROCEEDINGS{Wilamowski01analgorithm,
    author = {Bogdan M. Wilamowski and Okyay Kaynak},
    title = {An algorithm for fast convergence in training neural networks},
    booktitle = {Proceedings of the International Joint Conference on Neural Networks, 2:1778–1782},
    year = {2001}
}

Bookmark

OpenURL

Abstract

In this work, two modifications on Levenberg-Marquardt algorithm for feedforward neural networks are studied. One modification is made on performance index, while the other one is on calculating gradient information. The modified algorithm gives a better convergence rate compared to the standard Levenberg-Marquard (LM) method and is less computationally intensive and requires less memory. The performance of the algorithm has been checked on several example problems. 1

Citations

2304	Learning Internal Representations by Error Propagation - Rumelhart, Hinton, et al. - 1986
583	Learning representations by back-propagating errors - Rumelhart, Hinton, et al. - 1986
283	Increased rates of convergence through learning rate adaptation - Jacobs
192	Training feed-forward networks with the Marquardt algorithm - Hagan, Menhaj
108	Second-Order Methods for Learning: Between Steepest Descent and Newton's Method - Battiti, First- - 1992
49	Accelerated Learning in Layered Neural Networks - Solla, Levin, et al. - 1988
40	Improving the convergence of the backpropagation algorithm. Neural Networks - Ooyen, Nienhuis - 1992
33	Backpropagation past and future - Werbos - 1988
31	Speed up learning and network optimization with extended back propagation - Sperduti, Starita - 1993
15	Conjugate gradient algorithm for efficient training of artificial neural networks - Charalambous - 1992
12	A cost function for internal representation - Krogh, Thorbergsson, et al. - 1990
11	Enhanced training algorithms, and integrated training/architecture selection for multilayer perceptron networks - Bello - 1992
10	Acceleration of back-propagation through learning rate and momentum adaptation - Miniani, Williams - 1990
9	MEKA - A fast, local algorithm for training feed forward neural networks - Shah, Palmieri - 1990
8	Back-propagation improvements based on heuristic arguments - Samad - 1990
5	Improving Convergence of Backpropagation by Handling Flat-spots in the Output Layer - Balakrishnan - 1992
4	A Modified Regression Algorithm for Fast One Layer Neural Network Training - Andersen, Wilamowski
2	Introducing Stochastic Process within the Backpropagation Algorithm for Improved Convergence”, presented at ANNIE'94 - Salvetti, Wilamowski
2	An Empirical Comparison of FlatSpot Elimination Techniques in Back-Propagation Networks - Parekh, Balakrishnan, et al. - 1992
2	Modification of the Backpropagation Algorithm for Faster Convergence”, presented at - Torvik, Wilamowski - 1993