Scholarlink

The increasing use of technology in modern-day online learning environments has prompted educational institutions across the globe to focus on the concept of ‘student engagement’. This concept underlines the prediction of behavioral, cognitive, and emotional involvement of students in the learning process. The classification and prediction of student engagement in virtual scenarios is important for educational institutions because improved student engagement is believed to be a catalyst for improved student academic performance. In view of the significance of student engagement prediction, several researchers have put forth various Machine Learning (ML) and Deep Learning (DL) techniques—such as Decision Trees, Support Vector Machine, Convolutional Neural Network, Logistic Regression, and Random Forest—to classify student engagement levels. This study proposes an innovative ensembled approach, involving the combined use of Convolutional Neural Network (CNN) and Extreme Learning Machine (ELM) techniques, for student engagement classification. The proposed CNN-ELM approach, when applied to two publicly available datasets—FER-2013 and RAF-DB—marks a notable enhancement over the previous techniques because it is a fast and more accurate technique for classifying student engagement in technology-assisted education systems.

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

Using New Ensembled Approach to Identify Student Engagement Levels During Learning Process

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Kudratdeep Aulakh,
  • Rajendra Kumar Roul,
  • Manisha Kaushal

摘要

The increasing use of technology in modern-day online learning environments has prompted educational institutions across the globe to focus on the concept of ‘student engagement’. This concept underlines the prediction of behavioral, cognitive, and emotional involvement of students in the learning process. The classification and prediction of student engagement in virtual scenarios is important for educational institutions because improved student engagement is believed to be a catalyst for improved student academic performance. In view of the significance of student engagement prediction, several researchers have put forth various Machine Learning (ML) and Deep Learning (DL) techniques—such as Decision Trees, Support Vector Machine, Convolutional Neural Network, Logistic Regression, and Random Forest—to classify student engagement levels. This study proposes an innovative ensembled approach, involving the combined use of Convolutional Neural Network (CNN) and Extreme Learning Machine (ELM) techniques, for student engagement classification. The proposed CNN-ELM approach, when applied to two publicly available datasets—FER-2013 and RAF-DB—marks a notable enhancement over the previous techniques because it is a fast and more accurate technique for classifying student engagement in technology-assisted education systems.