UCL Discovery

Abstract

One way to understand meta-heuristic optimisation process is to look at it as an automated trial-and-error problem solving technique. This view provides interesting parallels to video gaming, which often requires the gamers to solve various problems using trial-and-error approaches. This paper considers the possibility of formulating optimisation problems in SBSE as video games and discusses challenges that are anticipated. The paper concludes with a demonstration of a proof-of-concept implementation that casts test data generation problem as a spatial puzzle.