상세정보
학술대회
Comparison of AI Techniques for Fighting Action Games - Genetic Algorithms/Neural Networks/Evolutionary Neural Networks
- 발행일
- 200709
- 출처
- International Conference of Entertainment Computing (ICEC) 2007 (LNCS 4740), v.4740, pp.55-65
- 협약과제
- 07MC1500, 멀티코아 CPU 및 MPU기반 크롯플랫폼 게임기술 개발, 양광호
- 초록
- Recently many studies have attempted to implement intelligent characters for fighting action games. They used genetic algorithms, neural networks, and evolutionary neural networks to create intelligent characters. This study quantitatively compared the performance of these three AI techniques in the same game and experimental environments, and analyzed the results of experiments. As a result, neural network and evolutionary neural network showed excellent performance in the final convergence score ratio while evolutionary neural network and genetic algorithms showed excellent performance in convergence speed. In conclusion, evolutionary neural network which showed excellent results in both the final convergence score ratio and the convergence score is most appropriate AI technique for fighting action games. © IFIP International Federation for Information Processing 2007.
- KSP 제안 키워드
- AI techniques, Action Games, Evolutionary neural networks, Genetic Algorithm, International federation, Score ratio, convergence speed, excellent performance, information processing