A Comparative Analysis of Decision Tree Classification Algorithms for Blended Learning Analytics in WEKA

Abstract

This study explores the application of decision tree classification algorithms for analyzing student performance data within a blended learning environment. The analysis, conducted using WEKA 3.8.6, focused on four attributes believed to influence student performance: course type, course level outcome (CLO), topic learning outcome (TLO), and level of assessment. A comparative analysis of J48, Random Forest, and SimpleCart algorithms revealed valuable insights. J48 demonstrated efficiency in model building, while Random Forest offered a balance between interpretability and accuracy. SimpleCart achieved the highest classification accuracy but could be less interpretable. The selection of the optimal algorithm depends on the analytical goals. J48 is suitable for rapid exploration, while SimpleCart prioritizes accuracy. Random Forest offers a compromise for scenarios where both understanding and accuracy are important. This study provides a foundation for understanding student performance through decision trees and highlights opportunities for further exploration using additional attributes, rule-based learners, and other machine learning algorithms. By leveraging these techniques, educators within blended learning environments can gain a deeper understanding of student performance and tailor their practices to optimize learning outcomes.