The model for assessing the risk of depression of students with data mining techniques.

Abstract

The purposes of this research are 1) to study and develop a risk techniques model for depression among students that may cause self-harm using data mining and the sample used in the research is 600 undergraduate students, Nakhon Pathom Rajabhat University randomly sampling from all faculties. The research instruments are 1) general information questionnaires and 2) PHQ-9 depression test assessment. The data analysis techniques used in this research consist of five data mining techniques: Random Tree, LMT, PART J48, and JRIP, the efficiency is compared by a Cross-validation test.

The results show that the data using the Random Tree is the most efficient with an accuracy of 96.00%, followed by the LMT with an accuracy of 95.90%, PART with an accuracy of 95.10%, J48 with an accuracy of 94.80%, and JRIP with an accuracy of 94.00% respectively.