Predicting Early Withdrawal of University Students: A Comparative Study between KNN and Decision Tree

Abstract

“The rising trend of students dropping out of universities without completing their degrees is becoming a concerning issue for institutions. To address this problem, the reasons behind this phenomenon need to be explored. However, most educational data sets have small sample sizes and varying patterns. Currently, there are few machine learning approaches for Pakistani higher education student performance. This study presents a machine learning-based approach to predict student withdrawals and identify the reasons behind them. The proposed approach compares two supervised ML algorithms, K-N earestNeighbors (KNN) and Decision-Tree (DT). The most important attributes affecting student retention are also determined using the ExtraTreesClassifier ensemble learning algorithm. In our experimental evaluation, the accuracy of KNN was 75%, and 70% for DT.”