Analysis And Prediction Of Community Crime Rate Based On Data Mining

The community is the basic unit of people's life,which is closely related to everyone's life.Nowadays,how to carry on the early warning analysis of community crime is of great significance.At present,the public security system and related industries have accumulated a large amount of data,which has created the prerequisite for data mining.In this paper,for the purpose of community crime rate analysis and prediction,data mining method has been used,combining community data and related crime data,establish the community crime rate analysis and forecasting system,reveals the inherent regularity between community crime rate and its influencing factors,obtains the key factors related to the community crime rate,and predicts the potential community crime rate,provides the scientific basis for the police to deploy the limited resources and the key investigation.The main contents of this paper are as follows:1.The establishment of data mining based on community crime rate analysis and forecasting system.The k-means clustering method is used to divide the community into three categories:high,middle and low crime rates,horizontal comparison of the attributes of communities with different crime rates,and through the ridge regression and lasso regression,and finally use the support vector machine to predict the potential crime rate of the community,provide technical and theoretical support for the reasonable evaluation of the security of the community.2.Multi-community data and crime data processing.The community and crime data were pretreated by Lagrangian interpolation,z-score data normalization method and principal component analysis(PCA)for the large and diversified characteristics of community data and related crime data,in order to create a prerequisite for the excavation.3.The application of community crime rate analysis and forecasting system.In order to verify the feasibility of the community crime rate analysis and forecasting system proposed in this paper,combined with a set of real community data and related criminal data to achieve the crime rate analysis and prediction system analysis,and achieved a better classification and prediction effect,points out the safety index of the community by using the evaluation system proposed in this paper,which can be used to evaluate the potential crime rate of the community.4.According to the data of community crime,capital income and the number of police,setting up the relationship between the three parts,and the equilibrium solution is obtained by experiment and quantitative distribution fitting,community policing resources and provide decision support for crime prevention work.