Time Series Analysis Of Theft

In the police practice,the public security organs have accumulated a lot of criminal intelligence data and criminal information.But only with the crime of data is not enough,it can not play a role in police practice,it is important that the need for mathematical principles and the use of specialized technical and statistical methods to improve the comprehensive assessment of criminal data,and according to the specific time,a specific area of a specific crime law and prevention and control to provide decision-making basis.There are some problems about the methods of crime data analysis among the scholars in contemporary China.Firstly,the deficiency that data supporting for descriptive analysis about theft crime still exists.Secondly,the predictive crime analysis is lack of practicability,such as the enthusiasm for analyzing the big trend of crime instead of practice policing data.Thirdly,the methods of analysis are too simple that the quantitative results lack accuracy.For instance,mathematical modeling and perspective scale cannot satisfy the needs of practice.This study aims at the exploring the patterns of theft in X district of H city and predicting the crime trends in near future base on time series analysis of police calling data with the assistance of statistics and computer science,only that could propose the crime prevention strategies combines with theories of Sociology Criminology.As we all concerned,it is practical and essential for the assistance of crime prevention among departments of Public Security.With the analysis of police recording data from 110 system of public security department in X district,H city,the records of theft considerably constitute 78% of police records in total and show the significant patterns in the time series.Therefore,this study has completed the analysis of time patterns and the modeling mathematical of predicting the trends in 2016 base the theft police calling data over the period of five years,that assisted by the EXCEL,SPSS22.0 and EVIEWS 5.0.All the outcomes are for the aim for the assistance of crime prevention among departments of Public Security.Firstly,the methods for analysis consist of frequency analysis,mean value analysis,central tendency analysis and cluster analysis.As a result,there are significant patterns among different units of time(day,week,month,year),but different patterns exist among various types of theft.Secondly,the outcomes of time series analysis.The degree of fitting of Exponential Smoothing Model is higher than ARIMAR(Autoregressive Integrated Moving Average)after comparing analysis.Obviously,the predictive analysis is based on the ESM.Lastly,proposing the microcosmic crime prevention strategies combined the analysis outcome and the theories of anomie,social study and sub-cultural.