Using decision trees to predict the likelihood of high school students enrolling for university studies

Abstract

This article presents the use of decision trees to identify the main factors which predict the likelihood of high school students matriculating at the Department of Geoinformatics, Palacky University in Olomouc (Czech Republic). The Department of Geoinformatics has been running a continuous and systematic information campaign about studying the fields of geoinformatics and geography within the department. In order to collect feedback about the information campaign, students who apply to study at the department are then given a questionnaire. Answers received from this questionnaire in two years, (2016 and 2017), were analyzed using decision trees that help us understand what specific type of information positively affects the likelihood of a student actually commencing studies at our department.