Development of a Genetic Algorithm Model: GAP
Chan Wai, Nelson
csnchan@comp.polyu.edu.hk
Department of Computing
Hong Kong Polytechnic University
Hung Hom, Hong Kong.
Abstract
Genetic Algorithms(GA) are quite popular in the past twenty years. Genetic Algorithms could be applied to a wide range of applications such as cost analysis, market timing, investment strategies, scheduling and etc. In general, genetic algorithms are simple, application-independent and can quickly perform a search to locate a near-optimal solution. To enable students to use GA for problem-solving, a simple learning tool is desired. The learning tool must be of low cost, simple to use, and above all should consist of various GA features. This paper proposes the architecture of a GA model for computer-assisted-learning. GAP is an acronym that stands for Genetic Algorithms Platform. GAP allows student to input his/her own evaluation function, population size, number of generation and etc. Standard cross-over methods are provided by GAP such as one-point, two-point and uniform cross-over. In addition, GAP can accept student defined cross-over method and produce the fittest member. GAP is built with certain intelligence so that it can regulate the convergence direction in case students supply inappropriate parameters. A prototype has been developed for the purpose of studying the feasibility and operations of the GAP model. The study and development of the prototype is a project funded by the Educational Development Group of the Hong Kong Polytechnic University.
© Asian Technology Conference in Mathematics, 1998. |
|
Go Back |
