A new class of quasi-Newtonian methods for optimal learning in MLP-networks

Authors:
A. Bortoletti;C. Di Fiore;S. Fanelli;P. Zellini
Affiliations:
Dipt. di Matematica, Univ. di Roma "Tor Vergata", Rome, Italy;-;-;-
Venue:
IEEE Transactions on Neural Networks
Year:
2003

Citing 0
Cited 8

Structured matrices in unconstrained minimization methods

Contemporary mathematics
A Hybrid Training Algorithm for Feedforward Neural Networks

Neural Processing Letters
Low complexity secant quasi-Newton minimization algorithms for nonconvex functions

Journal of Computational and Applied Mathematics
Minimization of a Detail-Preserving Regularization Functional for Impulse Noise Removal

Journal of Mathematical Imaging and Vision
Approximation BFGS methods for nonlinear image restoration

Journal of Computational and Applied Mathematics
Self-scaled conjugate gradient training algorithms

Neurocomputing
A two-stage algorithm integrating genetic algorithm and modified Newton method for neural network training in engineering systems

Expert Systems with Applications: An International Journal
Review article: A self-organizing map-based initialization for hybrid training of feedforward neural networks

Applied Soft Computing

Quantified Score

Hi-index	0.01

Visualization

Abstract

In this paper, we present a new class of quasi-Newton methods for an effective learning in large multilayer perceptron (MLP)-networks. The algorithms introduced in this work, named LQN, utilize an iterative scheme of a generalized BFGS-type method, involving a suitable family of matrix algebras L. The main advantages of these innovative methods are based upon the fact that they have an O(nlogn) complexity per step and that they require O(n) memory allocations. Numerical experiences, performed on a set of standard benchmarks of MLP-networks, show the competitivity of the LQN methods, especially for large values of n.