The University of Edinburgh -
Division of Informatics
Forrest Hill & 80 South Bridge

MSc Thesis #9742

Title:User-Guided Evolutionary Algorithms
Date: 1997
Abstract:Genetic Algorithms and other evolutionary algorithms mimic natural or biological evolution in many respects. Like nature, they encode potential solutions to a problem into artificial structures called chromosomes, they work blindly to find ever fitter individuals. the evolution process works automatically, starting off from a random point till a specific criterion is satisfied. In some situations, it is desirable to allow a user (an expert or novice) to direct the evolutionary process toward the user's own desired problem solutions. This is for instance, the case when we have imprecise problems that cannot be expressed specifically or mathematically. The solutions to such problems are based on human intuition and aesthetic tastes. This project seeks to investigate this approach to artificial evolution processes with a view to implementing a user-guided genetic algorithm for an aspect of interior designing.

[Search These Pages] [DAI Home Page] [Comment]