A Model for Reducing Indicator System of Evaluation of Public Security Condition Based on Rough Sets

摘要

The indicator system of evaluation of public security condition has many problems such as stronger subjectivity, more redundant information etc. So this paper proposed a model for reducing indicator system of evaluation of public security condition based on rough sets. The model firstly does data discretization in indicator system of evaluation of public security condition using the method of attribute discretization based on Shannon entropy. Then the model does attribute reduction based on the theory of rough sets to delete the redundant indicator in the indicator system. The model finally computes the weight of all remainder indicators. The analysis of the realistic example shows that this model is effective and feasible.