The Constructing Algorithm of Classification Knowledge Model Based on Information Entropy
Using equivalence relations to classify objects in rough sets, two objects are classified into different equivalence classes if they have a different attribute value. If the attribute of classification excessive and the attribute value couldnt describe the object exactly, then these objects are classified into different equivalence classes even they have same characteristic in reality, thus the result is not ideal. According to cognition psychology, the classification knowledge is acquainted with high certain attribute value, and low uncertain attribute value is neglected. The concept that people obtain is common characteristic of same kind of things. Using characteristics that information entropy can measure uncertain information, and choose high certain property, then produce equivalence classes; at the same time, using the structure of concept lattice, the classified knowledge model is constructed. According to the experiment result, construction algorithm of classified knowledge model is put forward in this paper, which is valid and feasible.
Abstract: Using equivalence relations to classify objects in rough sets two objects are classified into different equivalence classes if they have a different attribute value. If the attribute of classification excessive and the attribute value could
HAN Xie ZHANG Yuan YANG Xiaowen
Department of Computer Science and Technology, North University of China, Taiyuan 030051
国际会议
北京
英文
2007-08-05（万方平台首次上网日期，不代表论文的发表时间）