4.7. Merge ECG and EEG Features for Stress Detection

This step involves combining EEG and ECG features with participant metadata to create a unified dataset for stress classification. The process is illustrated in the following Figure 9.

Figure 9: Workflow for Merging and Processing EEG and ECG Data for Stress Detection

The merged dataset undergoes filtering to select specific emotions—anger, fear, and calmness—used to label stress and non-stress states. A binary label is then assigned, with anger and fear categorized as stress (1) and calmness as non-stress (0). Next, data types are standardized, and any missing or infinite values are handled to ensure data consistency.