EEG-based cognitive state classification using a leakage-safe stacked ensemble learning approach

Key Points

• Cognitive state classification improves accuracy through a stacked ensemble learning approach, and could enhance brain-computer interfaces.
• The stacked ensemble method achieves over 90% classification accuracy, highlighting its effective application in EEG data analysis.
• Assessment using advanced machine learning techniques demonstrates enhanced performance in classifying cognitive states from EEG data compared to traditional methods.
• Supports the development of more reliable brain-computer interfaces, though further validation across diverse populations is necessary.