A Predictive Analytics Approach in Determining the Predictors of Student Attrition in the Higher Education Institutions in the Philippines

Abstract

The paper identified the predictors of student attrition in the Higher Education Institution (HEI) through predictive analytics approach. The prediction model used in the study includes variable optimization through Genetic Algorithm (GA) and decision tree generation phase through C4.5 algorithm. The college student leavers' data from one of the Higher Education in the Philippines from the school year 2008-2009 until the school year 2018-2019 was used as datasets of the study. Out of forty identified reasons for leaving as variables, there were nine (9) identified predictors of student attrition. Through the identified predictors, administrators of educational institutions may design intervention plans related to the student attrition.