A Comparative Study of Different Machine Learning Techniques to Predict the Result of an Individual Student Using Previous Performances

Abstract—Machine learning is a sub-field of computer science
refers to a system’s ability to automatically learn from experience
and predict new things using the learned knowledge. Different
machine learning techniques can be used to predict the result
of the students in examination using previous data. Machine
learning models can recognize vulnerable students who are at
risk and take early action to prevent them from failure. Here,
a model was developed based on the academic performance
of the students and their result in the SSC exam. This paper
also shows a comparative study of different machine learning
techniques for predicting student results. Five different machine
learning techniques were used to demonstrate the proposed
work. They are Naive Bayes, K-nearest Neighbours, Support
Vector Machine, XG-boost, Multi-layer Perceptron. Data were
preprocessed before fitting into these classifiers. Among the five
classifiers, MLP achieved the highest accuracy of 86.25%. Other
classifiers also achieved a satisfactory result as all of them were
above 80% accuracy. The results showed the effectiveness of
machine learning techniques to predict the performance of the
students.
Index Terms—Machine learning, Result, Prediction