Knowledge structure preserving fuzzy attribute reduction in fuzzy formal context
International Journal of Approximate Reasoning(2019)
摘要
Attribute reduction is an important and even crucial pre-processing step in knowledge achievement. Existing fuzzy attribute reduction methods in Formal Concept Analysis tend to remove the redundant attributes that preserve the invariability of algebraic structures, and thus might distort the knowledge structures underlying fuzzy formal contexts. To avoid the issue, this paper presents a fuzzy attribute reduction approach of preserving the invariability of knowledge structure, and discusses some properties of fuzzy attribute reduct under the choice of hedges. We also propose an algorithm for computing fuzzy attribute reduct using Łukasiewicz adjoint operators and hedge. Compared with the previous approaches, our approach is based on the widely accepted framework, fuzzy concept lattice with hedge, and has the capability of producing a vector, specifying a significance degree for each attribute.
更多查看译文
关键词
Fuzzy formal context,Fuzzy attribute implication,Fuzzy attribute reduction,Feature selection
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络