New globally convergent training scheme based on the resilient propagation algorithm

In this paper, a new globally convergent modification of the Resilient Propagation-Rprop algorithm is presented. This new addition to the Rprop family of methods builds on a mathematical framework for the convergence analysis that ensures that the adaptive local learning rates of the Rprop's sc...

Full description

Saved in:
Bibliographic Details
Published in:Neurocomputing (Amsterdam) 2005-03, Vol.64, p.253-270
Main Authors: Anastasiadis, Aristoklis D., Magoulas, George D., Vrahatis, Michael N.
Format: Article
Language:English
Subjects:
Batch learning
Convergence analysis
First-order training algorithms
Global convergence property
IRprop
Rprop
Supervised learning
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, a new globally convergent modification of the Resilient Propagation-Rprop algorithm is presented. This new addition to the Rprop family of methods builds on a mathematical framework for the convergence analysis that ensures that the adaptive local learning rates of the Rprop's schedule generate a descent search direction at each iteration. Simulation results in six problems of the PROBEN1 benchmark collection show that the globally convergent modification of the Rprop algorithm exhibits improved learning speed, and compares favorably against the original Rprop and the Improved Rprop, a recently proposed Rrpop modification.
ISSN:0925-2312
1872-8286
DOI:10.1016/j.neucom.2004.11.016