Research on Influence Diagrams Evaluation with Sampling and Genetic Algorithms
Influence Diagrams (IDs) is a graphical knowledge representation of decision problem,it may be viewed as an extension of Bayesian Networks,with additional node types for decision and utilities. It is widely used in many fields such as decision analysis and uncertainty reasoning problems. Evaluating an influence diagrams means finding an optimal decision for each of its decision nodes. A global solution to the decision problem consists of a series of decisions that maximizes the expected utility. Evaluation algorithm plays a vital role for the application of Influence diagram. This paper studies the sampling algorithm and genetic algorithm to evaluate influence diagram,and make experiments of these two evaluation algorithms,analyze and compare the performances of the two algorithms.
Influence Diagrams evaluation Sampling Algorithm Genetic Algorithm
Chen Jie Zhou Lihua Yang Ming Liu Lingliang
Department of Computer Science,School of Information,Yunnan University,Kunming 650091,China Yunnan University of Traditional Chinese Medicine,Kunming 650500,China
国际会议
南宁
英文
1-4
2010-12-10（万方平台首次上网日期，不代表论文的发表时间）