🧠 EEG Emotion Classifier

A machine learning pipeline for emotion classification from EEG signals.
Currently implemented with classical ML methods (SVM) on the preprocessed DEAP dataset.

📂 Project Structure

eeg-emotion-classifier/
├── data/ # raw and processed EEG data (not included in repo)
├── src/ # preprocessing, feature extraction, models, pipeline
├── results/ # evaluation results (confusion matrix, ROC curve, metrics)
├── main.py # entry point
├── requirements.txt # dependencies
└── .gitignore

⚙️ Installation

Clone the repository and install the dependencies:

git clone https://github.com/goktug-sirma/eeg-emotion-classifier.git
cd eeg-emotion-classifier
pip install -r requirements.txt

▶️ Usage

1. Download the DEAP dataset

   • Required files: data_preprocessed_python.zip
   • Extract the archive and place all sXX.dat files into data/raw/.

2. Run the pipeline:

python main.py

📊 Results

Results will be added after running on real data. Expected outputs:

• results/confusion_matrix.png
• results/roc_curve.png
• results/metrics.txt

🗺️ Roadmap

• ✅ Project structure and pipeline skeleton
• ✅ Preprocessing and feature extraction
• ✅ SVM training and evaluation
• ⬜ Add results with real datasets
• ⬜ Explore cross-subject generalization
• ⬜ Experiment with deep learning models (CNN, LSTM, Transformers)

📜 License

This project is licensed under the MIT License.