Application of Naive Bayes to Students’ Performance Classification

Abstract

Naive Bayes Classifier is a strong tool or model in classifying students' performance based on various factors. Thus, this research developed a classification model that can accurately classify students into different academic performance categories. The study utilized data, collected from 1,422 students at the University of Ibadan, Nigeria. Descriptive statistics and data visualization techniques were used to gain insights into the distribution and relationships among the variables. Subsequently, a Naive Bayes classifier model was built using 70% of the data for training and 30% for testing. In addition, a Support Vector Machine (SVM) model was built to compare with the performance of the Naive Bayes model. The results of the descriptive statistics show that the respondents comprise of 846 females and 576 males. From the female respondents, 144 of them had First Class grade, 432 had Second Class Upper, 252 had Second Class Lower, and the remaining 18 had Third Class. From the male respondents, 144 of them had First Class grade, 198 had Second Class Upper, 216 had Second Class Lower, and the remaining 18 had Third Class. The Naive bayes model achieved an overall accuracy of 87%, while the SVM model achieved an overall accuracy of 85%. The results highlighted that department, grade in the first year, and monthly allowance were the most crucial features for classifying performance outcomes, while gender, age group and whether or not the respondents’ parents are educated, exerted the least significant influence on the models. Thus, on average, the Naive Bayes model outperformed the SVM in the classification of students’ performance based on the data collected. Also, the early academic performance, and financial support are significant factors in determining students' overall performance in the Institution.