Estimates on weight-decay regularization by variable-basis schemes

Authors:
Giorgio Gnecco;Marcello Sanguineti
Affiliations:
University of Genoa, Department of Computer and Information Science, Genova, Italy;University of Genoa, Department of Communications, Computer, and System Sciences, Genova, Italy
Venue:
ACS'09 Proceedings of the 9th WSEAS international conference on Applied computer science
Year:
2009

Citing 3
Cited 0

Learning with generalization capability by kernel methods of bounded complexity

Journal of Complexity
The sample complexity of pattern classification with neural networks: the size of the weights is more important than the size of the network

IEEE Transactions on Information Theory
Sequential greedy approximation for certain convex optimization problems

IEEE Transactions on Information Theory

Quantified Score

Hi-index	0.00

Visualization

Abstract

The optimization problems associated with supervised learning from data by the weight-decay and related techniques are described. Suboptimal solutions expressed by connectionistic models with a given number of computational units are investigated. In the final part of the paper, improvements of some estimates obtained in [1] for the same optimization problems are given.