Adapting to a Realistic Decision Maker: Experiments towards a Reactive Multi-objective Optimizer

The interactive decision making (IDM) methods exploit the preference information from the decision maker during the optimization task to guide the search towards favourite solutions. This work measures the impact of inaccurate and contradictory preference information on the quality of the solutions generated by the IDM methods. The investigation is done in the context of the BC-EMO algorithm, a recently proposed multi-objective genetic algorithm.

• Type: Conference
• Year: 2010
• Where: LION
• Authors: Paolo Campigotto, Andrea Passerini
• Keywords: Contradictory Preference Information, LION 2010, Multi-objective Genetic Algorithm, Optimization, Preference Information