项目名称: 基于证据理论的多源模糊信息系统的不确定性研究
项目编号: No.61300121
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 自动化技术、计算机技术
项目作者: 冯涛
作者单位: 河北科技大学
项目金额: 27万元
中文摘要: 不确定性广泛存在于现实世界中,其表现形式多种多样,可用信息熵、模糊性、粗糙性等方法来度量。粒计算是研究粒化思维方式的一种理论,从其观点来看,不确定性是知识在不同知识粒度层次上的某种表现形式。证据理论作为经典概率的弱化结果,给出了信任结构和多源信息系统的融合方法,在处理多源不确定信息融合、推理和决策规则获取方面拥有独特的优势。本项目拟将粒计算、证据理论相结合,借鉴模糊粗糙集和已有的研究不确定性的理论与技术,系统地研究基于证据理论的多源模糊信息系统的不确定性问题。主要内容包括:(1)基于粒计算的多源模糊信息系统的知识不确定性度量;(2)多源模糊信息系统的证据的不确定性;(3)多源模糊信息系统的决策规则的不确定性。本项目不但能够丰富证据理论的研究成果,为多源模糊信息系统的数据挖掘及不确定性分析提供新的理论与方法,而且在规则提取、机器学习等方面有广泛的应用价值。
中文关键词: 粒计算;不确定性;证据理论;模糊粗糙集;规则合成
英文摘要: Uncertainty is widely existing in the world, which has a variety of forms. It can be measured by using the method of information entropy, fuzziness, roughness, etc. Granular computing is a theory of studying granular thinking. From the view of granular computing, uncertainty is some expression form of information at different information granular layers. As a weakened result of the classical probability, the evidence theory gives belief structure and the fusion method of multi-information systems, and has a unique advantage in dealing with multi-source uncertain information fusion, reasoning, and making decision rules. This project is planning to combine granular computing and the evidence theory, using fuzzy rough sets and the existing theories and technologies of uncertainty, then systematically study the uncertainty of multi-fuzzy information systems based on the evidence theory. The main objects are: (1) the uncertainty measure of knowledge in multi-fuzzy information systems based on graular computing; (2) the uncertainty of evidence in multi-fuzzy information systems; (3) the uncertainty of decision rules in multi- fuzzy information systems. The research results of this project will not only enrich the evidence theory, providing new theories and methods of data mining and uncertainty analysis of multi-fuzz
英文关键词: granular computing;uncertainty;evidence theory;fuzzy rough sets;rule fusion