Comparison of Evolutionary Algorithms and their Hybrids Applied to MarioAI

Researchers have been applying artificial/ computational intelligence (AI/CI) methods to computer games. In this research field, further researchesare required to compare AI/CI methods with respect to each game application. In thispaper, we report our experimental result on the comparison of evolution strategy, genetic algorithm and their hybrids, applied to evolving controller agents for MarioAI. GA revealed its advantage in our experiment, whereas the expected ability of ES in exploiting (fine-tuning) solutions was not clearly observed. The blend crossover operator and the mutation operator of GA might contribute well to explore the vast search space.

• Hidehiko Okada
• Yuki Fujii

• Evolutionary algorithm
• autonomous game controller agent
• neuroevolutions
• MarioAI

[1] H.-P.Schwefel, Evolution and Optimum Seeking. New York: Wiley & Sons, 1995.
[2] D. E.Goldberg, Genetic Algorithms in Search Optimization and Machine
Learning. Addison Wesley, 1989.
[3] X. Yao, "A review of evolutionary artificial neural
networks,"International Journal of Intelligent Systems, vol. 4, pp. 539-567, 1993.
[4] K.O. Stanley and R. Miikkulainen,"Evolving neural networks through
augmenting topologies, "EvolutionaryComputation, vol.10, no.2,
pp.99-127, 2002.
[5] L.J.Eshelman, "Real-coded genetic algorithms and
interval-schemata,"Foundations of Genetic Algorithms 2, pp.187-202,1993.
[6] Y.S. Ong, M.H.Lim, N. Zhu and K.W.Wong, "Classification of adaptive
memetic algorithms: acomparative study,"IEEE Transactions on Systems
Man and Cybernetics-Part B, vol.36, no.1, pp.141-152, 2006.
[7] J.E.Smith, "Coevolving memetic algorithms: areview and progress
report,"IEEE Transactions on Systems Man and Cybernetics -Part B,
vol.37, no.1, pp.6-17, 2007.
[8] F. Neri, C. Cotta, and P. Moscato (eds), Handbook of Memetic
Algorithms. Springer, 2011.