Student academic performance prediction using transfer learning

Department: Computer Science
By:

jaymeg

Project ID: 8610

Rating:

(5.0) votes: 1

Rate this project

Price:₦5000

Abstract

Predicting student academic performance is crucial for identifying at-risk students and enhancing educational outcomes. Traditional methods often rely on limited data and predefined models, which may fail to generalize across diverse learning environments. To address these challenges, this research explores the use of transfer learning for predicting student academic performance. Transfer learning allows the leveraging of pre-trained models from related domains, adapting them to predict student outcomes with greater accuracy and efficiency. In this study, a transfer learning-based approach is applied to a dataset containing various student attributes such as previous academic scores, study habits, extracurricular activities, and personal factors. The model fine-tunes a pre-trained neural network to recognize patterns that contribute to academic success, while minimizing the data requirements typically associated with building predictive models from scratch. Evaluation of the model is conducted using key performance metrics, including accuracy, precision, recall, and F1-score, to assess its ability to correctly predict high and low-performing students. Results demonstrate that transfer learning significantly improves the accuracy of academic performance predictions compared to traditional machine learning techniques. Additionally, this approach provides better generalization across different student cohorts and educational settings....

Preview +

Other Computer Science project topics and materials you might be interested in

Student academic performance prediction using transfer learning

Upload Project

Departments