Determining Student Demographic Attributes Influencing Performance Using Binary Classification in KDP Model

Abstract

Abstract Machine learning (ML) is one way that can help decipher the intricate relationship between students' data and their performance. When implemented correctly in learning environments, machine learning will improve knowledge of fundamental processes by simplifying the identification, extraction, and evaluation of underlying factors that affect student learning and levels of achievement. This study employed the experimental research approach using binary classification techniques based on the six-step Knowledge Discovery Process (KDP) model. Five classifiers were used within the Rapid Miner's 9.10.010 educational environment as both experimental and analytical tool. The dataset comprised of 2334 records, 17 attributes with one class variable (students’ semester average score) inclusive. Twenty different tests were conducted. The experiments' results were evaluated using 10-fold cross-validation and ratio split validation with bootstrap sampling. The Random Forest algorithm (RF), Rule Induction methods (RI), Naive Bayes (NB), Logistic Regression (LR) and Deep Learning (DL) algorithms were used in the experiment. The experimental results demonstrated that the RF method outperforms the other four techniques in all six-evaluation metrics that were employed for the selection process with the accuracy being 93.96%. According to the RF classifier model, the mother's and father's education levels of students are two recognized demographic factors per this study that significantly influence pre-tertiary students’ academic achievement. This study has significantly reduced the gap in practical knowledge observed in the literature by introducing an intervention scheme for respective student's requiring intensive or minimal academic interventions in its prediction procedure.