摘要
本文给出处理二级(或阶)不确定性的一种内涵方法(或称为基于模型的方法),该方法用概率簇空间表示证据空间的二级不确定性结构,用L-集合表示规则强度的二级不确定性,用概率簇空间的传递方式刻画二级不确定性的传播过程.给出了概率簇上限定化关系的传递性质,并用限定化的概念对规则组合的两种不同方式进行了比较.本文为基于知识的推理提供了新的综合框架.
This paper provides a model-based rnethod for processing the second-level uncertainties. In this paper, the authors use the space of probability class to present the second-level uncertainty structure of evidence space, L-set to present the second-level uncertainties of the rule strengths and use transmitting way of spaces of probability classes to describe the transmitting process of the second-level uncertainties. In this paper, the authors also present some transmitting properties of the specialization relations on classes of probabilities, and compares the two different ways of rule combination by using the concept of specialization. This paper gives a new cornprehensive framework for knowledge-based reasoning.
出处
《软件学报》
EI
CSCD
北大核心
1997年第7期535-540,共6页
Journal of Software
基金
国家863高科技项目
国家自然科学基金
国家教委博士点基金
国家教委符号计算与知识工程开放实验室的资助
关键词
二级不确定性
概率簇空间
限定化
专家系统
The second-level uncertainties, the space of probability class, specialization