A Survey on Student Academic Performance using Machine Learning

Abstract

The students have to plan and set goals to achieve best performance in their engineering career. So to obtain good results throughout the year it is essential for them to improve and maintain their academic result. This academic performance helps them to get good jobs as well as to get good colleges for further studies. Student academic performance also helps the institution to keep a track of the performance of their student’s progress in their academics and also helps to analyze the students that are weaker in their academics. Various data mining techniques are used to extract knowledge from the academic data which will help the students to know their weakness and can improve their result. In this study, every student’s CGPA of previous semesters is taken into consideration and various classification algorithms are used to classify and predict their CGPA in their next semesters. These classification algorithms will be compared by building student academic performance prediction model based on their previous semesters CGPA and other conditions using WEKA tool. Some classification algorithms like Naive Bayes, K-Nearest Neighbor and Random Forest are used and each of the algorithm’s average accuracy and error rate is determined and then the best classification algorithm will be selected to get better results. The records of student marks of all branches Computer, EXTC and IT in Xavier Institute of Engineering College will be taken to build the prediction model.