Dissertation > Industrial Technology > Automation technology,computer technology > Automated basic theory > Artificial intelligence theory

Research on Rough Sets Theory and Its Applications in Knowledge Acquisition

Author MaYuLiang
Tutor ZhaoGuangZhou
School Zhejiang University
Course Control Theory and Control Engineering
Keywords Rough Sets Theory (RST) Knowledge acquisition Roughness of knowledge Attribute Importance Evaluation Attribute discretization Rule Reduction Value Reduction Information Fusion
Type PhD thesis
Year 2005
Downloads 760
Quotes 11
Download Dissertation

Pawlak's rough set (Rough Sets, abbreviated RS) theory is a process of knowledge, particularly imprecise, incomplete knowledge of a new mathematical tool. The theory of knowledge gives a formal definition, making the knowledge to carry out effective analysis and operations. In addition, RS theory also provides a set of data from the automatic knowledge acquisition tool, the reduction of knowledge. Currently, RS theory is being widely used in artificial intelligence, pattern recognition, and many other fields. Based on the characteristics of rough set theory, will be applied to the field of knowledge acquisition, knowledge acquisition can support multiple steps, such as data preprocessing, data reduction, rule generation, data access and other dependencies. Main text as follows: The first chapter is summarized rough set theory, including the following four parts: the basic concepts of rough set theory; rough set theory is different from other smart features; rough set theory commonly used application software; rough set theory research directions, including theoretical and applied research. Then get an overview of knowledge, and finally thesis describes the content and structure of the paper. The second chapter introduces knowledge information entropy and mutual information based on the concept, rough set theory are discussed in detail roughness of knowledge and information in the relationship between the angle of knowledge from the information given in a quantitative roughness characterization. Then from the information theory point of view of the main concepts of rough set gives a new representation, which we call information representation, and its intuitive meaning and rationality described. Chapter III presents information based on rough set theory importance of evaluation system properties. Using rough set theory under and upper approximation concept definition an assessment parameter α_R (X), and thus determine the size of the parameter importance of system properties. With the existing evaluation system parameter method, a method based on rough set theory can get more information about the system. The fourth chapter introduces the information system attribute discretization significance, steps, classification, and several existing methods. Then use the decision table compatibility feedback, we propose a domain independent continuous attributes division algorithm. Finally the algorithm and there are already several ways to do a comparative analysis, obtained satisfactory results. Chapter V of the information system reduction algorithm. First studied using rough set theory to real-valued information system attribute reduction method, discrete and attribute reduction method to study the relationship, the real value of the property proposed a discrete method of automatically determining attribute category, combined with rough set theory gives information systems for real-valued attributes reduction algorithm; then combine probabilistic method for fuzzy sets has been studied, and use phase

Related Dissertations
More Dissertations