基于退火演化算法的知识获取机制的研究*

Research on Annealing-Genetic Algorithm Based Knowledge Acquisition

中文关键词  知识获取  组合优化  退火演化算法  

英文关键词  knowledge acquisition  combinatorial optimization  annealing-genetic algorithm

中文摘要

      针对专家系统知识获取的“瓶颈”问题， 本文从优化角度建立了一个最终可被归结为大规模组合优化问题的知识获取模型， 并通过基因遗传算法和模拟退火算法结合后所构成的退火演化算法求解该模型， 从而实现知识库的自动生成.

英文摘要

      Knowledge acquisition is a “bottleneck” for building expert system.ln this paper,a realvalued criterion is proposed to evaluate the performance of the rule. On this basis,knowledge acquisition is reduced to a large-scale combinatorial optimization problem. A hybrid approach with the genetic algorithm being em- bedded in simulated annealing,which is defined as annealing-genetic algorithm,is proposed to solve the prob- lem. By this strategy,the knowledge base can be automatically generated from examples.