Student Dropout Analysis with Application of Data Mining Methods

One of the indicators of potential problems in the higher education system may be a large number of student dropouts in the junior years. An analysis of the existing transaction data provides the information on students that will allow the definition of the key processes that have to be adapted in order to enhance the efficiency of studying. To understand better the problem of dropouts, the data are processed by the application of data mining methods: logistic regression, decision trees and neural networks. The models are built according to the SEMMA methodology and then compared to select the one which best predicts the student dropout. This paper concentrates primarily to the application of the data mining method in area of higher education, in which such methods have not been applied yet. In addition, a model, useful for strategic planning of additional mechanisms to improve the efficiency of studying, is also suggested.