Attribute Reduction Based On Fuzzy Inner Product Dependence And Discernibility Matrix

As a new mathematical tool, the classical rough set model is mainly used to deal with uncertainty information, the theory of data analysis, which was founded in 1982 by Poland scientist Z Pawlak. They are very powerful in dealing with large capacity data. Rough set theory is very important, very valuable theoretical knowledge. Rough set not only establishes a strict mathematical model, it is applied in data mining, artificial intelligence, decision analysis and other fields.The general idea of this paper is to analyze the problem of the classical fuzzy rough set dependence function. Based on the knowledge of the original fuzzy rough set, the fuzzy product dependency and fuzzy discernibility matrix attribute reduction model is constructed.The data and experimental analysis of the model are carried out. The main work of this paper is as follows:1. Firstly, analysis of the deficiency of the classical fuzzy rough set dependence function, based on the knowledge of the original fuzzy rough set, the upper and lower approximation sets of fuzzy rough sets based on distance functions are defined,we construct the inner product fuzzy decision dependence function and study the monotonicity properties.And the algorithm is designed.The experimental results show that the algorithm is effective.2. Secondly, brief introduction of the basic knowledge of neighborhood discernibility matrix. Analysis of the deficiency of the classical fuzzy rough set dependence function. New fuzzy identification matrix model is constructed, and the new identification degree is introduced. A new attribute reduction algorithm is proposed.Experimental results show that the algorithm can find a smaller and more efficient feature subset. It is feasible.