Scholarlink

This research examines the influence of emotional states specifically positive (happy) and negative (sad) emotions, on students’ cognitive performance during learning activities using EEG signals. Emotions significantly influence attention, problem-solving, reasoning and memory. We focused on 20 subjects aged 21–30 using the Allegers Virgo 40-channel EEG device. Preprocessing involved filtering and ICA, followed by time and frequency domain feature extraction. FFT analysis revealed differences in brain activity patterns influenced by emotional states and gender. Different classification tasks involved the utilization of several Machine Learning Models such as Logistic Regression, SVM, Gradient Boosting, Random Forest, and Stacking classifiers. From which Extreme Gradient Boosting (XGB) and Random forest gives accuracy 90.91, 90.54% whereas gradient boosting, and Stacking classifier gives highest accuracy Such as 93.24% with respect to these happy and sad class. The study highlights the substantial effect of emotions on cognitive functions like attention, problem-solving, reasoning, and memory, offering valuable insights to enhance educational practices and interventions.

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

Classification of the Emotional States for Analysis of the Cognitive Aspect of Student During Learning Activities Using Machine-Learning Techniques

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Pratibha R. Bhise,
  • Sonali B. Kulkarni,
  • Talal A. Aldhaheri

摘要

This research examines the influence of emotional states specifically positive (happy) and negative (sad) emotions, on students’ cognitive performance during learning activities using EEG signals. Emotions significantly influence attention, problem-solving, reasoning and memory. We focused on 20 subjects aged 21–30 using the Allegers Virgo 40-channel EEG device. Preprocessing involved filtering and ICA, followed by time and frequency domain feature extraction. FFT analysis revealed differences in brain activity patterns influenced by emotional states and gender. Different classification tasks involved the utilization of several Machine Learning Models such as Logistic Regression, SVM, Gradient Boosting, Random Forest, and Stacking classifiers. From which Extreme Gradient Boosting (XGB) and Random forest gives accuracy 90.91, 90.54% whereas gradient boosting, and Stacking classifier gives highest accuracy Such as 93.24% with respect to these happy and sad class. The study highlights the substantial effect of emotions on cognitive functions like attention, problem-solving, reasoning, and memory, offering valuable insights to enhance educational practices and interventions.