A Monte Carlo-Based Model for Predicting Student Performance

Accurate prediction of student academic performance distributions is essential for institutional planning and decision support. This paper introduces a novel Monte Carlo (MC) framework for forecasting cohort-level grade distributions using historical data. The MC model extracts empirical probability distributions and generates predictions via stochastic sampling, explicitly capturing uncertainty and variability in student outcomes. Evaluated across engineering, medical, and aggregated student cohorts, the MC model outperformed a linear regression benchmark, particularly in capturing nonlinear grade transitions and stochastic variability. The MC approach preserves distributional structure and offers superior fidelity in modeling cohort behavior, serving as a reliable, uncertainty-aware complement to traditional machine learning. This work presents a validated probabilistic methodology and evaluation framework for predicting academic performance.