Hybrid Fuzzy Logic-based Explainable Models for Assessing Student Academic Outcomes

In this study, a unique method for predicting students' grades by combining Fuzzy Logic with Particle Swarm Optimization (PSO) is presented. The Student Performance dataset is employed by the algorithm to forecast the final grade (G3) based on the previous grades (G1, G2). The effectiveness of the proposed system is demonstrated through a comprehensive model evaluation that employs significant metrics, including R-squared, Mean Absolute Error, Root Mean Squared Error, and Mean Squared Error. In contrast to conventional regression methodologies, the hybrid model exhibits superior performance, with MSE values of 0.345, RMSE values of 0.588, MAE values of 0.482, and R2 values of 0.944. These findings indicate that Fuzzy Logic and PSO have the potential to improve regression models that predict the performance of students in the classroom. The proposed method is a reliable for educational data analytics due to its superior predictive accuracy compared to linear regression and other baseline models.