Visualization for Genetic Algorithms

Abstract

Information visualization is still evolving alongside the incredible technological advances of the past decades as we can now use computers to process exceptionally large datasets easily and represent them through animated and interactive visualizations. The development of this field has been largely dependent on its work with other fields by finding new ways to visualize data and finding meaningful patterns of information. One of these fields is artificial intelligence, which often turns to nature for inspiration in problem-solving and devises tools such as the genetic algorithm, a search heuristic which simulates natural selection in order to find and optimize solutions to particular problems. In this dissertation we covered some of the most important developments from the fields of information visualization and genetic algorithms, and detail the process of the creation of a new visualization tool. This takes the form of a functional prototype which can process the data acquired from a genetic algorithm and, using visualization techniques which had not been applied to this particular field before, is able to effectively represent meaningful patterns in the data which lead to significant conclusions.