Research And Implementation Of Community Correction Staff Risk Assessment Technology

With the continuous advancement of relevant policies in China,it has also promoted various research and applications of various machine learning algorithms in the historical data of prisoners in the community correction have been carried out.How to use historical data to achieve the risk assessment of community corrections personnel is the most practical one.Based on the real data of community corrections,this paper proposes a risk assessment technique based random forest-based for community corrections.The assessment includes a classification forecast for community corrections personnel during the correction period.The risk assessment technology is applied to the Ministry of Justice community correction information management system.The main research contents are as follows:(1)In view of the actual data of the current community corrections personnel data,such as large amount of data,disorderly and irregular,and sensitive information involved,the pre-treatment is pre-processed.The specific process is as follows:1.Data feature screening and sensitive data culling,data discretization conversion and data correction.2.Using the OOB(out-of-bag data)feature selection on the community corrections data to obtain the data sets with the greatest correlation.(2)This paper proposes a random forest-based risk assessment technique for community correctors and implements it in an actual production environment.By using historical data to establish a classification prediction model,it is predicted whether new community corrections personnel will be out of control(re-crime)during the correction period.By using the random equilibrium sampling method to obtain training data for the category imbalance in the data set,and using the random forest to obtain the risk assessment model.The application of the evaluation model in the Java production environment is implemented using PMML.(3)Design and implement a community corrections risk assessment module integrating data collection,data statistics and data mining in the national community correction information management system to demonstrate the effect.According to the historical data of community correction,the risk assessment model was tested and analyzed,and the performance and practical effects of the community corrections risk assessment technology proposed in this paper were verified.modeling.