Student Graduation Prediction Using Algoritma C 4.5 With Fitur Selection Chi Square

Abstract

Prediction study period is required by the university management in determining the policy of preventive related to prevention since the beginning of the case DO (Drop Out) This study aims to determine the factors of academic influence on these students can graduate on time or not, and build predictive models best with data techniques .Criteria elections mining model used is the chi-square method selection feature. C4.5 algorithms generated by that period of study is influenced by indeks prestasi of semester, the number of repeated courses, courses taken and the shots certain subjects. Therefore factors - these factors can be used as an evaluation for the manager of the college.
Keyword: Chi Square, C 4.5, prediction, students,datamining