Predicting Cognitive Workload in Flight Simulations using EEG Spectral and Connectivity Features

Note: The publication link for the paper is not available yet.

Research Overview

In this research, we aim to investigate the predictive potential of frontal-parietal connectivity in determining cognitive workload during flight simulation. We utilize EEG connectivity features to analyze the brain's electrical activity and its relationship to cognitive workload levels.

Code and Features

We employ Support Vector Machines (SVMs) for predictive modeling, and we utilize the feature selection technique, Recursive Feature Selection (RFE), to identify the most important features. The feature set includes both Phase Locking Value (PLV) features, and (relative) Spectral Power features.

Repository Structure

• raw/: Stores the datasets used in this study. The raw dataset is not available on GitHub.
• data/: Stores the feature sets per participant in subdirectories (VR & Desktop) as .npz files. The dataset is not available on GitHub.
• code/: This directory contains the code used for data preprocessing, feature selection, SVM modeling, and evaluation.
• results/: This directory contains plots obtained from the experiments.

Citation

If you find this work or the code provided in this repository useful for your research, please consider citing our paper once it becomes available. The citation details will be provided in the paper itself.

Contact

For any questions or inquiries regarding this repository, please reach out to us via email at bas.verkennis.code@gmail.com.