🧠 EEG Neuroheadset Eye State Prediction

🔍 Project Overview

This project develops a machine learning model to predict whether a person’s eyes are open or closed based on EEG (electroencephalography) signals recorded from 14 different brain regions.

🎯 Objectives

• Process EEG data to extract meaningful insights.
• Compare multiple machine learning models to determine the best classifier.
• Evaluate performance using accuracy and other relevant metrics.

📌 Dataset: Includes 14 EEG features with a manually labeled eye state ('0' for open, '1' for closed).

🧹 Data Preprocessing

• Standardization: Used StandardScaler to normalize EEG signals for better model performance.
• Train-Test Split: Split dataset into 80% training and 20% testing for fair model evaluation.
• Class Balance Check: Verified that the dataset was relatively balanced (~55% closed, ~45% open).

🔬 Model Comparisons & Performance

✅ Logistic Regression

• Insights: Struggled to capture complex EEG patterns due to its linear nature.
• Performance: Limited effectiveness in distinguishing between eye states.

🌳 Random Forest Classifier

• Insights: Demonstrated strong predictive accuracy with EEG data.
• Performance Metrics:
  • Accuracy: High classification accuracy.
  • Feature Importance: Certain EEG channels were more influential in predicting eye state.

🤖 Neural Networks (Future Work)

• Potential improvement using deep learning models such as LSTMs or CNNs for EEG signal classification.

📊 Tech Stack

• Python
• Pandas & NumPy (Data manipulation & preprocessing)
• Scikit-learn (Machine learning models)
• Matplotlib & Seaborn (Data visualization)

📌 How to Use

1. Clone this repository:

   git clone https://github.com/YOUR_USERNAME/eeg-neuroheadset.git

2. Install dependencies: bash - pip install pandas numpy scikit-learn matplotlib seaborn
3. Open and run the Jupyter Notebook to explore the analysis.

📜 License

This project is open-source and available under the MIT License.