Reduction Algorithm For Research And Application Based On Variable Precision Rough Set

As information technology and database technology developing rapidly, people every day face the enormous amount of data,Data mining is a technology that dedicated to data analysis and understanding, revealing hidden knowledge of the internal data ,and is currently a very active area of research of AI. Rough set theory is an effective way of dealing with ambiguity and uncertainty of the mathematical tools for data mining research has provided new ideas and the foundation.This paper studies the variable precision rough set reduction algorithm ,for traditional data mining deal with the noise problem of insufficient data, from both theoretical and applied aspects of reduction algorithm in-depth study.Main functions include:1) Re-interpret the concept of the classic rough set based on the variable precision rough set theory; analysis of rough set theory in data mining applications, the theoretical basis and rationale, and point out research directions.2) A comparative analysis of two kinds of variable precision rough set model of the reduction algorithm, namely, theβlower approximation andβlower distribution reduction algorithm, combining the two algorithms proposed an improved algorithm and verify that the new algorithm.3) Propose an assessment model based on variable precision rough set and entropy, and the model was applied to evaluation of enterprise independent innovation capacity, through empirical analysis confirms the model capability of independent innovation in the enterprise evaluation of effectiveness.