Drug Traffickers' Attribute Association Analysis And Probability Prediction

Today,with the development of all aspects of society,the criminal means of committing are more diversified,therefore more scientific methods should be used in the work of police officers.The probability prediction of drug traffickers and the correlation analysis of drug traffickers' properties means that the theory and method are used to analyze and judge the criminal offense,and provide the basis for the prevention and control of crime.The progress of science and technology makes the database widely used,police departments have accumulated some data,However,the arrival of the information age makes the amount of data continues to increase,these data are only stored in the system,coupled with the special field of public security,there are still a lot of problems in the analysis of intelligence,and there is little research on it.In this thesis,the association rules of data mining are applied to the information management of drug traffickers,trying to find some valuable rules and information,and we will also use the scientific methods to make reasonable prediction,the police can eventually guide the handling of police arrangements to improve the efficiency of law enforcement and rapid response capability,timely prevention and fight against crime.This thesis uses the data provided by the police department.In the real problem,the ordinary Apriori algorithm does not take into account the different records have different importance,so there are some limitations,this thesis make some improvements and find the property which is more close to the drug trafficking;The logistic model is used to predict the probability of the crime of drug traffickers,and the results are validated by 10-fold cross validation.