Activities to Predict Students’ Final Grades Based on Machine Learning Techniques
Activities to Predict Students’ Final Grades Based on Machine Learning Techniques

Summary/Abstract: The paper analyses a student performance dataset using the classification model. This approach can help improve the academic performance of struggling students by implementing specific procedures and strategies before exams, aiming to enhance educational attainment. To build a classification and ensemble models and predict student performance, a dataset with 120 instances was used, each of which has 39 attributes, with an implementation of NaiveBayes, Decision Tree, Neural network, k-Nearest Neighbors, and Support Vector Machine algorithms. This study's results offer valuable insights into student performance assessment and highlights the importance of data mining in education, specifically through diverse evaluation methods such as the base classifier, attribute selection, and AdaBoost meta-classification. High accuracy was obtained after applying the meta-classification method.

• Details
• Contents