Comprehensive Analysis of Students’ Performance by Applying Machine Learning Techniques

Predicting the performance of students becomes more interesting due to the increase in the number of students opting for engineering courses and parents’ inclination to see their wards becoming engineers and getting placed in outstanding companies with good packages. This paper aims at performance analysis of students by applying different machine learning techniques. The collected data contains the students’ personal information, family background, friends, study time and other information which contains 45 questions. Various classifiers like Random Forest, SVM and Logical Regression were applied on the collected data by hard-coding in Python using Jupyter Notebook. Based on the classification algorithms, statistics are generated and comparison of all three classifiers is made so as to envisage the truthfulness and to discover the finest acting algorithm. Random Forest classifier gives 96% accuracy, Logistic Regression gives 86% accuracy and SVM classifier gives 84% accuracy. So, the best result is shown by Random Forest classifier. © 2020, Springer Nature Singapore Pte Ltd.