Study On Attribute Reduction Of Incomplete Neighborhood Data With Variable Precision

The extraction of valid data from a large amount of data requires the data to be processed.The introduction of rough set theory has brought important breakthroughs in dealing with large-scale data and uncertainties.In the information system containing large-scale data,the data loss will be caused when the classical rough set theory is used to reduce the continuous data.The neighborhood rough set model can effectively solve this problem.At this stage,neighborhood rough set model is widely used in dealing with mixed data of numbers and symbols.However,the existing neighborhood rough set calculation methods mostly study the complete mixed data,and discuss the attribute reduction under the incomplete mixed data relatively.In this paper,the data is static and dynamic two cases,for the missing,continuous,symbolic data such as information systems,respectively,designed the corresponding data processing methods,the following brief:When the data is static,an information system containing missing,continuous,symbolic and other data is first combined with the multi-granularity rough set to analyze the upper and lower approximation operators of the incomplete neighbor decision-making system under the variable precision model And attribute reduction.Secondly,a variable precision rough set model for generalized neighborhoods is constructed by neighborhood granulation,and an evaluation method of attribute importance is given.Based on this,an incomplete neighborhood decision is designed System attribute reduction algorithm.Finally,the validity of the algorithm is verified by an example analysis.In real life,the large amount of complicated data in many application fields often changes dynamically,and the dynamic reduction of the data will affect the original attribute reduction results.Aimed at incomplete mixed numerical and symbolic data,that is,incomplete neighborhood decision-making system,an evaluation method of importance degree of measurement attribute is put forward first,and the change of conditional entropy caused by incremental update of object is analyzed in detail Situation and updating mechanism;Secondly,an attribute-reduction incremental updating method based on variable-precision attribute reduction is constructed,which can directly deal with complex data with incomplete numeric and symbolic coexistence.Finally,through the example the results show that the algorithm can solve the attribute reduction results of incomplete neighborhood data with variable accuracy.