Dropout prediction system to reduce discontinue study rate of information technology students

• Published in: 2018 5th International Conference on Business and Industrial Research (ICBIR) pp. 110 - 114
• Main Authors: Limsathitwong, Kittinan, Tiwatthanont, Kanda, Yatsungnoen, Tanasin
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.05.2018
• Subjects: Business; classification; Data mining; Data models; Decision trees; dropout prediction models; educational data mining; Information technology; machine learning algorithms; Prediction algorithms; Predictive models; web application
• DOI: 10.1109/ICBIR.2018.8391176

Nowadays, The student dropout is a serious problem which has an impact on education in Thailand. This research is a case study of the students at Thai-Nichi Institute of Technology between first-year and second-year that have a high rate retirement. There are analyze in order to find dropout rate and develop the web application to predict status of the students from their grade of each subjects. The prediction models were developed to use Decision Tree algorithms as well as Random Forest Algorithms to achieve an improvement. The Decision Tree classifier has obtained precision, recall and F1-Measure of 0.80, 0.92 and 0.85, respectively. Results show that overall results of the predictor are satisfactory. The application can recognize dropout students and identify those students who need special attention that is very useful to help the students improving their learning process and to monitor the student performance in a systematic way.