Comparison of learning rules for adaptive population-based incremental learning algorithms

Abstract:

This paper describes the adaptive approach of the Population-based Incremental Learning (PBIL) algorithm, and proposes several Learning Rules aimed to improve its performance. The assessment of such alternatives was made in terms of both the convergence time and of the quality of the achieved solutions. Two classical optimization problems were used for the tests: The Job Shop Scheduling problem and the Traveling Salesman problem. The obtained results are very promising and suggest that some of the proposed learning rules have a superior performance, without degrading drastically the quality of the solutions.

Año de publicación:

2012

Keywords:

• Optimization
• Adaptive algorithms
• Learning rules
• Incremental learning

Fuente:

scopus