Prediction of student's academic achievements by using the data mining methods

For an eligible education, higher education institutions are to conclude the right decision by means of administrative and educational aspect. For academic planning being missing or incorrect, students being unsuccessful, determining the roadmap for graduates, student?s dropping out are the subject of higher education institution?s problem. Solving these problems and taking measures are very important for eligible education. There is a bulk of increasing data belonging to education of higher education institutions. These increasing data has nothing to do with administration, academician and education. These data can be meaningful by processing with the methods of data mining, estimations can be done with higher accuracy rate and measures can be taken. Data mining methods are powerful tools for academic interventions. In this thesis, several prediction techniques in data mining such as artificial neural networks and Decision tree methods are used to help the educational institutions to predict the students? graduation scores. In this context, 127 unique student records of Computer Education and Instructional Technology department are used, where the students were in the university between 2006-2010 and 2007-2011. Two different scenarios are investigated. Firstly, we used the students? first two years scores. Secondly, students? first three years scores are used for prediction. According to the experimental results, it is observed that the ANN yielded better performance than the Decision tree and the second scenario yielded better results than the first scenario.Key Words: Educational Data Mining, Prediction of Student?s Academic Achieve-ments, Artificial Neural Networks, Decision Trees.