Machine Learning Models for Student Performance Prediction

• Published in: 2023 International Conference on Innovative Data Communication Technologies and Application (ICIDCA) pp. 27 - 32
• Main Author: Pande, Shilpa Mangesh
• Format: Conference Proceeding
• Language: English
• Published: IEEE 14-03-2023
• Subjects: Correlation; Data Visualization; Education; K Nearest Neighbors; Kernel; Logistic Regression; Machine learning; Predictive models; Sociology; Student performace system; Support Vector Machine; Support vector machines
• DOI: 10.1109/ICIDCA56705.2023.10099503

Education is the key to success which provides multiple opportunities in the life. Education not only contributes to individual success but also contributes to the success of the society. The Indian education system still follows the traditional way of the teaching learning process which lacks in interactive sessions. So, it becomes very difficult to continuously monitor student performance. Indian population is huge which makes the monitoring and analyzing student performance challenging. The Indian education system lacks in the standards for accessing the student performance and achievements. The proper methodology is required to monitor student's academic progress and development. Multiple parameters influence a student performance so it is a very challenging task to find out the significant parameters which are affecting a student performance. Another challenge in the education system is to identify the slow learners at an early stage. In this research work, various factors affecting student performance are analyzed and visualized like student's age, parent's education & occupation, health etc. The visualization technique is used to identify the weak students at an early stage to work on their improvement. To forecast the student performance, a variety of Machine Learning (ML) methods, including K Nearest Neighbors (KNN), Logistic Regression, and Support Vector Machine (SVM), are used. The SVM model with linear kernel gave the best accuracy 84.37%for the selected dataset.