A Machine Learning Based DSS in Predicting Undergraduate Freshmen Enrolment in a Philippine University

TL;DR

This paper develops a logistic regression-based decision support system to predict undergraduate freshmen enrollment in a Philippine university, aiding institutions in enrollment planning amid educational reforms.

Contribution

It introduces a predictive model using logistic regression integrated into a user-friendly DSS for enrollment forecasting in Philippine higher education.

Findings

The model achieved high accuracy in predicting enrollment likelihood.

Key factors influencing enrollment were identified.

The DSS provides real-time predictions and visualizations.

Abstract

The sudden change in the landscape of Philippine education, including the implementation of K to 12 program, Higher Education institutions, have been struggling in attracting freshmen applicants coupled with difficulties in projecting incoming enrollees. Private HEIs Enrolment target directly impacts success factors of Higher Education Institutions. A review of the various characteristics of freshman applicants influencing their admission status at a Philippine university were included in this study. The dataset used was obtained from the Admissions Office of the University via an online form which was circulated to all prospective applicants. Using Logistic Regression, a predictive model was developed to determine the likelihood that an enrolled student would seek enrolment in the institution or not based on both students and institution's characteristics. The LR Model was used as the algorithm in the development of the Decision Support System. Weka was utilized on selection of features and building the LR model. The DSS was coded and designed using R Studio and R Shiny which includes data visualization and individual prediction.