Scholarlink

This paper analyses a dataset from the Mohammed V University of Rabat's MOOC platform to understand factors influencing the performance of learners. Preprocessing and analysis of a dataset of 29,994 learners revealed that factors like age, gender, quiz scores, and homework assignments are positively correlated with learner achievement. The study employed various machine learning algorithms such as k-Nearest Neighbors, Random Forest, Catboost, and Logistic Regression. The Catboost algorithm demonstrated superior performance in predicting learner success, achieving an accuracy of 86.1%. The model based on Catboost can potentially guides the identification of targeted interventions and improves the overall quality of online learning experiences.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Use of Catboost Algorithm to Enhance the Prediction of the Learner Performance in MOOCs

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Abdelkarim Taam,
  • Amine Amar,
  • Brahim Hmedna,
  • Khalid Benabbes,
  • El Menzhi Kaoutar,
  • Adil El Makrani

摘要

This paper analyses a dataset from the Mohammed V University of Rabat's MOOC platform to understand factors influencing the performance of learners. Preprocessing and analysis of a dataset of 29,994 learners revealed that factors like age, gender, quiz scores, and homework assignments are positively correlated with learner achievement. The study employed various machine learning algorithms such as k-Nearest Neighbors, Random Forest, Catboost, and Logistic Regression. The Catboost algorithm demonstrated superior performance in predicting learner success, achieving an accuracy of 86.1%. The model based on Catboost can potentially guides the identification of targeted interventions and improves the overall quality of online learning experiences.