Bataan Peninsula State University

This system uses machine learning models trained on historical academic data like grades, GWA, enrollment status (regular/irregular), gender distribution, drops, and INCs, to surface early trends and risk indicators that manual review alone can miss. Student records are referenced only by an anonymized student ID, never by name, so no individual identity is exposed on the dashboard. Model reliability varies by prediction type, and that reliability is shown transparently on the dashboard rather than presented as one certain number, so results are meant to inform judgment, not replace it. Models are retrained as new academic data is uploaded each term, keeping forecasts current without overstating how certain any forecast of the future can be. By design, this system is built for decision support, flagging at-risk records or cohorts who may benefit from additional academic assistance, and is not intended as a performance score for any college, department, or dean, so the incentive stays on early intervention rather than on managing the numbers themselves.