Ensemble voting machine learning model for prediction of campus placement of the student

Placements in campus interviews are the dream of every student in college. Placement in campus interviews is a vital measure of an educational institution’s standards and student performance. Machine learning with the knowledge discovery process helps to forecast the student’s performance in on-campus interviews. This paper suggested an ensemble model-based voting classifier with BayesNet and J48 is used to classify the student’s academic data and forecast the placement opportunity. This work compares two ensemble stacking models and a voting-based classification model with J48 for obtaining an efficient model of placement prediction. Both ensemble stacking models use BayesNet and J48 classifiers as the base classifiers. The J48 classifier is used as the meta classifier in one stacking process and the voted perceptron is used in another. In the ensemble voting model, BayesNet and J48 are used as the base classifiers and the probability average of a class of base classifiers is used for the combination rule. The ensemble voting model gains high accuracy with a minimum error rate than other models. This model produced 91% of accuracy in the placement prediction. J48 and BayesNet classifiers are combined with probability average-based combination rules in the ensemble voting model.