Recommending Graduate Admission Using Ensemble Model

Potential scholars are often faced with the challenge of picking which institutions to submit applications to for degree courses. Although a lot of forecasters and advisors are available to help students, decisions are based on only a few instances of prior enrolment. Thus, they can occasionally be accurate. The majority of prognosis algorithms in use today were created using machine learning (ML) techniques. We provide a stacking ensemble framework to forecast and assess educational achievement in courses of study. To increase precision, three techniques for ML are constructed in the preliminary stage: Decision Tree (DT), K-Nearest Neighbour (KNN), and Naïve Bayes (NB). Logistic regression (LR) is subsequently added through stacking. For every applicant, the algorithm generates a recommendation for entrance to the institution of their choice based on an algorithmic approach that takes into account their interests and educational experiences. We built the computational framework to leverage information from a real-time admissions dataset of 500 grads. Recall, accuracy, precision, and F1-score are the 4 assessment indicators used in the present study. Comparative investigations show that the suggested model outperforms the individual base models. The evaluation’s key factors may be interpreted and leveraged as a guide for choosing possible candidates for enrolment.