Mathematical Modelling of Latent Variables of Students' Performance of Public Engineering Universities in Mathematics Using Machine Learning

Abstract

This research paper provides a mathematical modelling and analysis for students’ performance in mathematics in public engineering universities of Sindh Pakistan. The research distinguishes and measures the effect of different latent variables such as Institute Environment, Mathematics Anxiety, University Reputation (Students’ Opinion) and University Facilities all these latent variables have been found from 36 observed variables and Proposed model is designed. Additionally, Random Forest (RF), Support Vector Machine (SVM), Gradient Boosting Machine (GBM) and Artificial Neural Networks (ANN) are used for comparison and validation of proposed model of machine learning which is designed. Evaluation of these models have been done through Accuracy, precision, recall, F1 score, and MAE, MSE and RMSE have been used. The findings provide further evidence for the contribution of well-being, engagement, and prior success in mathematics achievement. The proposed model provides practical implications for decision-makers and educators to come up with focused policies and interventions to enhance mathematics learning performance of engineering students in Sindh.