Research Based On Data Mining Techniques And Its Application In Reform School Students

The number of reform school students in our country increased gradually,the research on correction of the students in the reform school made important progress in recent years.This thesis through the rational use of data mining techniques,combined a reform school's actual situation,do the scientific analysis and research on reform school students.The main work includes:Firstly,introduces the concept of data mining,and the basic principles.Discuss the data mining as a process of search the hidden useful information or connections in a large number of data sets and its commonly used methods and modes and discuss its significant achievement in real life.Among this,focused on the two classic modes:Association and Classification,we discussed its basic concept,introduced several classical algorithm in this field.Highlighted one algorithm which based on the Bayes'theorem's decision tree,analysised its differences from the ordinary decision tree algorithm,emphasized its advantages.Next,using these methods described,this thesis analyzed the reform school students database,found some significant association rules,propose a Bayesian model which can analysis of missing attribute in the student data,and using this establish a Bayesian decision tree which completed the classification of students in the reform school.Finally,on the basis of research and analysis in the reform school's student database,developed the reform school students classification system.This system is used to store all the previous student data,generating new padding data algorithms and make new student classification.The research and experiment of this classification system is able to guide the reform school student classification and correct student's bad behavior.