论文部分内容阅读
不确定性推理是人工智能的关键问题。本文引进了包含度的概念以及包含度的生成方法,给出了包含度在专家系统中关于知识获取与推理的某些应用。 不确定性是人工智能中最活跃的研究领域,对于发展智能计算机有着重要意义。不确定性推理有定量方法、定性方法以及两者相结合的方法。对于定量方法,首先是与测度及不确定性信息相关。不同的测度与表示方法得到不同的不确定性推理,如包含有MYCIN不确定因子与主观贝叶斯推理的概率推理方法、证据推理方法、模糊推理方法、信息推理方法等。上述方法的共同特征是用概率测度、信息测度、似然测度、可能性测度、必然性测度给出假设的度量。不确定性推理的本质是在各种测度下对于包含关系给出一种估计。概括已有的不确定性推理使我们引进了包含度的概念,它足以表达各种不确定性推理的共同特征。
Uncertainty reasoning is a key issue of artificial intelligence. This paper introduces the concept of degree of inclusion and the generation method of degree of inclusion, and gives some applications of degree of inclusion in expert system about knowledge acquisition and reasoning. Uncertainty is the most active area of research in artificial intelligence, which is of great importance to the development of intelligent computers. Uncertainty reasoning has quantitative methods, qualitative methods and a combination of the two methods. For quantitative methods, the first is to measure and uncertainty information. Different measures and representation methods derive different uncertainties, such as probabilistic reasoning method, evidence reasoning method, fuzzy reasoning method and information reasoning method, which include MYCIN uncertain factors and subjective Bayesian inference. The common feature of the above method is to use the probability measure, the information measure, the likelihood measure, the likelihood measure, the inevitability measure gives the hypothesis measure. The essence of uncertainty reasoning is to give an estimate of the inclusion relation under various measures. Generalization of existing uncertain reasoning led us to introduce the notion of inclusiveness, which is sufficient to express the common features of various uncertain inferences.