TriPat‑XFE: a triangle pattern‑based explainable feature engineering framework for EEG classification.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Artificial Intelligence

Background:

• Electroencephalography (EEG) signals offer a cost-effective, non-invasive method for brain activity monitoring.
• Extracting meaningful features from EEG data is crucial for accurate analysis and interpretation.
• Current methods may lack interpretability or require significant computational resources.

Purpose of the Study:

• To introduce a novel feature extraction method, Triangle Pattern (TriPat), for EEG analysis.
• To propose a new explainable feature engineering (XFE) framework integrating TriPat.
• To achieve high-performance classification with interpretable outputs for EEG data.

Main Methods:

• Developed Triangle Pattern (TriPat) for accurate and explainable EEG feature extraction.
• Integrated TriPat into an XFE framework with CWINCA for feature selection and tkNN for classification.
• Utilized Directed Lobish (DLob) AI for generating symbolic explanations from selected features.

Main Results:

• Achieved over 90% classification accuracy on Artifact, Stress, and Psychosis EEG datasets.
• Generated connectome diagrams visualizing brain activity patterns.
• Demonstrated superior accuracy and lower computational cost compared to existing methods.

Conclusions:

• The TriPat-centric XFE framework provides a practical and interpretable solution for EEG analysis.
• The framework operates efficiently on standard hardware, requiring no GPUs.
• Offers a unified approach for high-performance classification and explainable AI in neuroscience.