Genetic algorithms and mixed integer linear programs for optimal strategies in a student's "sports" activity

This paper uses an entertaining student "sports" game to illustrate that GAs can be adapted to problems with uncertain properties and complexity. These problems can be solved easily through GAs within a few seconds. Contrary to this, using standard MILP techniques does not yield results in a reasonable time. Categories and Subject Description: I.2.8 Artificial Intelligence:Problem Solving and SearchHeuristics General Terms: Algorithms, Performance

• Type: Conference
• Year: 2006
• Where: GECCO
• Authors: Thomas Butter, Franz Rothlauf, Jörn Grahl, Tobias Hildenbrand, Jens Arndt
• Keywords: GECCO 2006, Optimization, SearchHeuristics General Terms, Standard Milp Techniques, Uncertain Properties