A Genetic Programming Approach for the Traffic Signal Control Problem with Epigenetic Modifications

This paper presents a proof-of-concept for an Epigenetics-based modification of Genetic Programming (GP). The modification is tested with a traffic signal control problem under dynamic traffic conditions. We describe the new algorithm and show first results. Experiments reveal that GP benefits from properties such as phenotype differentiation, memory consolidation within generations and environmentally-induced change in behavior provided by the epigenetic mechanism. The method can be extended to other dynamic environments.