A Genetics-Based Machine Learning Approach for Realtime Scheduling

Abstract

In this paper, we adopt a genetic-based machine learning (GBML) approach to a realtime scheduling problem in which several products are to be assigned to one of the buffers, and propose a method of generating and selecting rules for assigning each product to a desirable buffer. In applying the GBML, we use the Pitt approach, where the set of rules (rule-set) is represented symbolically as an individual of genetic algorithms, and the fitness of an individual is calculated based on the total cost required for transporting, operating and keeping of whole products. Through some computational experiments, the effectiveness and the possibility of our approach is investigated.