Research On The Clustering Method Of The Criminal Cases Based On The Multiple Uncertainty

In the emergency situation,the information which is described and recorded by victims,witnesses and criminal investigators has the features of randomness,fuzziness and grayness.It leads to that the information of criminal cases have the strong multiple uncertainty feature.Then come up with the clustering method of the criminal cases based on the multiple uncertainty,and research on the aspects including cases clustering index system construction,clustering attribute index evaluation,clustering feature index weight allocation and cases clustering method design.(1)From the perspective of multiple uncertainties,I define the connotation of the criminal cases clustering based on the multiple uncertainty,build the clustering index system of criminal cases and design the clustering process of criminal cases based on the multiple uncertainty;(2)According to the feature of multiple uncertain numbers,I deeply analyze the relationships and commonalities of probability numbers,fuzzy numbers,interval-valued numbers and grey numbers.Then I proposed the concept and characterization methods of general interval grey number so that I can unify these uncertain numbers.Based on the general interval grey number,I research on the basic operation,norm model and comparison of it,and proposed a new way to solve the problem of multiple attribute clustering index weight configuration based on multiple uncertainty;(3)Through the study for the historical clustering cases,deeply mine the clustering rules,establish the general grey similarity model,measure the distance of indexes and cases,and build the general interval grey number entropy weight allocation model based on the historical clustering case study to solve the weight of clustering indexes.It is a new method to solve the problems of multi-attribute clustering index weight configuration based on the multiple uncertainty.(4)Combining the traditional grey whitening weight function of classic,moderate measurement and lower measurement with the general interval grey numbers,I infer four whitening weight function of general interval grey numbers.Then design the clustering method of criminal cases based on general grey entropy weight allocation and propose a new technique for the multiple attribute clustering decision problems.Finally taking the criminal cases of AnHui for example,I validate the models and methods in this paper and the results meet the reality.It also proves the feasibility and effectiveness of the model.