1.School of Computer Science and Technology,Xuzhou Normal University,Xuzhou 221116 2.School of Management,Hefei University of Technology,Hefei 230009 3. School of Environment and Spatial Informatics,China University of Mining and Technology,Xuzhou 221008
Abstract:To keep user rationality is a key element in interactive evolutionary computation to converge to the global solution. The maximum generation must be designed appropriately to help user keep rationality. Firstly, three different kinds of definition of the maximum generation are proposed. Secondly, the methods to calculate the maximum generation for six kinds of fitness-assignment methods are given. Both theory analysis and experimental results show that the most-satisfactory-identified fitness-assignment and the scale fitness-assignment practically help user keep rationality in more generations. The research provides references to select appropriate fitness-assignment methods.
[1] Holland J H. Adaptation in Natural and Artificial Systems. Cambridge, USA: MIT Press, 1992 [2] Gong Dunwei, Yuan Jie, Ma Xiaoping. Interactive Genetic Algorithms with Large Population Size // Proc of the IEEE Congress on Evolutionary Computation. Hongkong, China, 2008: 1678-1685 [3] Hao Guosheng, Huang Yongqing, Zhang Yong, et al. Rational User-A Sufficient Condition for Global Convergence in Interactive Evolutionary Computation. Pattern Recognition and Artificial Intelligence, 2008, 21(4): 441-445 (in Chinese) (郝国生,黄永青,张 勇,等.理性用户——交互式进化计算全局收敛的一个充分条件.模式识别与人工智能, 2008, 21(4): 441-445) [4] Gong Dunwei, Yuan Jie. Impact of Individuals Fitness Expressions on Interactive Genetic Algorithms Performances // Proc of the Chinese Control and Decision Conference. Guilin, China, 2009: 2415-2420 [5] Gong Dunwei, Sun Xiaoyan, Yuan Jie. Interactive Genetic Algorithm with Individuals Uncertain Fitness // dos Santos W P, ed. Evolutionary Computation. Vienna, Austria: In-Tech Publisher, 2009 [6] Hu Jing. Interactive Genetic Algorithm with Image Information Retrieval. Master Dissertation. Hefei, China: University of Science and Technology of China. School of Computer Science and Technology, 2001 (in Chinese) (胡 静.用于图形图象信息检索的交互式遗传算法.硕士学位论文.合肥:中国科技大学.计算机科学与技术系, 2001) [7] Zhou Yong, Gong Dunwei, Hao Guosheng, et al. Neural Network Based Phase Estimation of Individual Fitness in Interactive Genetic Algorithm. Control and Decision, 2005, 20(2): 234-236,240 (in Chinese) (周 勇,巩敦卫,郝国生,等.交互式遗传算法基于NN的个体适应度分阶段估计.控制与决策, 2005, 20(2): 234-236,240) [8] Gong Dunwei, Hao Guosheng, Zhou Yong, et al. Interactive Genetic Algorithms with Multi-Population Adaptive Hierarchy and Their Application in Fashion Design. Applied Mathematics and Computation, 2007, 185(2): 1098-1108 [9] Takagi H, Kishi K. On-line Knowledge Embedding for an Interactive EC-Based Montage System // Proc of the International Conference on Knowledge-Based Intelligent Information Engineering Systems. Adelaide, Australia, 1999: 280-283 [10] Denis P, Philippe C, Thierry B, et al. Eye-Tracking Evolutionary Algorithm to Minimize User Fatigue in IEC Applied to Interactive One-Max Problem // Proc of the Genetic Evolutionary Computation Conference. London, UK, 2007: 2883-2886 [11] Gong Dunwei, Guo Guangsong, Lu Li, et al. Adaptive Interactive Genetic Algorithms with Interval Fitness of Evolutionary Individuals. Progress in Natural Science, 2008, 18(3): 359-365 [12] Sugimoto F, Yoneyama M. Robustness against Instability of Sensory Judgment in a Human Interface to Draw a Facial Image Using a Psychometrical Space Model // Proc of the IEEE International Conference on Multimedia and Expo. New York, USA, 2000, Ⅱ: 635-638 [13] Sugimoto F, Yoneyama M. Hybrid Fitness Assignment Strategy in IGA-A Method to Compose Fitness // Proc of the IEEE Workshop on Multimedia Signal Processing. Vingin Islands, USA, 2002: 284- 287 [14] Huang Yongqing, Lu Qing, Liang Changyong, et al. Interactive Multi-Agent Evolutionary Algorithm and Its Application. Journal of System Simulation, 2006, 18(7): 2030-2032,2055 (in Chinese) (黄永青, 陆 青,梁昌勇,等.交互式多智能体进化算法及其应用.系统仿真学报, 2006, 18(7): 2030-2032, 2055)