Suboptimal Decision Table Reduction Algorithm Based On Neighborhood Rough Model

Because of confidentiality or historical reasons,missing data often exists in collecting data on the Internet,so that the classic set theory can't processing in the data processing.In order to solve this problem,rough set,which is a mathematical tool,arises at the historic moment.Rough set can deal with the imprecise data set,and to high-dimensional data simplification is one of the important technology in the development of the rough set now.With the development of rough set,a lot of expansion on the rough set have been proposed.The concept of neighborhood rough set is a kind of expansion of rough set,it can directly handle continuity data sets,solving practical problems.In this paper,we studies attribute Reduction based on the rough neighborhood model.And we redesigned Ant Colony Suboptimal Decision Table Reduction Algorithm(Ant Colony Suboptimal Decision Table Reduction Algorithm,shorthand for ACSR)based on Ant Algorithm(Ant Colony Optimization,ACO)to solve the optimal value.This paper mainly do the following tasks:(1)First of all,through the introduction to the classical rough set,raises the concept of neighborhood rough set,thus redefined set on the upper and lower approximation and neighborhood decision system in neighborhood rough,such as concept.These determine the basis for later research.(2)Neighborhood rough set algorithm is introduced based on the classical rough set algorithm.According to the existing study on the algorithm of processing data,we release paper proposed reduction algorithm of dealing with the cube.(3)Through the analysis of the experiments on UCI data sets,the paper verify the algorithm is an effective algorithm to solve the cube,and the formula for solving? operator is given.(4)Finally,the paper summarize the proposed algorithms.