Prediction of students’ performance in high school by data mining classification techniques

Abstract

Evaluation and prediction of students’ performance in high school help to find important factors affecting students’ success in education and moreover they can have an important role in helping educational managers in improving the quality of schools. According to this fact that data mining science has always been a suitable technique to extract knowledge from data, this can be used for providing a good approach. This article tries to present superior models in predicting students’ performance. The mentioned data of this article are taken from 386 students of high schools in Bushehr province. In presented models best classification algorithms in educational data mining towards presenting superior predicting models have been provided. analyze and validation done on models shows the obtained results are precise and reliable. in this regard individual, environmental and educational factors affecting successful and unsuccessful students have been analyzed and according to them efficient models based on decision tree methods like c4.5 tree algorithm, support vector machine methods and logistic regression have been presented. The results can help managers of educational systems towards a correct educational planning an optimizing of educational processes in high schools