Model Construction Using ML for Prediction of Student Placement

Abstract

Abstract: “Model construction using ML for prediction of student placement” aims to predict the placement of a student using various performance metrics on the Machine Learning algorithms. Early prediction makes the institutional growth as well as the student to get placed. It helps the student to prepare all the company requirements at early stage and monitors the student performance. Existed work was done on the algorithms like Logistic Regression (LR), Support Vector Machine (SVM), Naive Bayes. In the proposed work to predict the student placement considered dataset and applied data preprocessing to make the data easier to train the model for prediction using Decision Tree (DT) and XG Boost along with the existing algorithms. Accuracies are calculated using different performance metrics like Accuracy and F1-score, Precision, Recall. The algorithm that worked with the best accuracy is SVM with 91%, and the LR and DT algorithms got 88% accuracy whereas Naïve Bayes got 86% and then the XG Boost stood last with an accuracy of 84%. We are able to make a decision which algorithm is better than other algorithms. Higher accuracy algorithm is mostly preferred to predict the student performance.