Exoplanet Detection and Characterization

An end-to-end machine learning pipeline for detecting and characterizing exoplanets from stellar light curves using NASA Kepler and TESS mission data.

Overview

This project implements a transit detection system that identifies periodic dips in stellar brightness indicative of planetary transits. The pipeline processes photometric time series data and classifies candidate signals as confirmed planets, false positives, or noise.

Data Sources

• Kepler Data Release 25: ~200,000 stellar light curves with confirmed planet labels
• TESS Full-Frame Images: Ongoing mission data via MAST
• Kepler TCE Table: Threshold Crossing Events with disposition labels
• Kaggle Subset: Cleaned dataset of ~5,000 stars for initial development

Project Structure

exoplanet-detection-ml/
├── data/                    # Data storage (gitignored)
│   ├── raw/                 # Original downloaded data
│   ├── processed/           # Cleaned and feature-engineered data
│   └── external/            # Third-party reference data
├── src/                     # Source code
│   ├── data/                # Data loading and preprocessing
│   ├── features/            # Feature engineering
│   ├── models/              # Model implementations
│   └── visualization/       # Plotting utilities
├── notebooks/               # Jupyter notebooks for exploration
├── tests/                   # Unit and integration tests
├── docs/                    # Documentation
└── configs/                 # Configuration files

Key Features

• Light curve preprocessing (detrending, normalization, outlier removal)
• Transit feature extraction (depth, duration, periodicity)
• Multiple classification approaches (Random Forest, CNN, LSTM)
• Model interpretability and uncertainty quantification
• Reproducible experiment tracking

Requirements

• Python 3.10+
• Lightkurve (NASA light curve analysis)
• Scikit-learn
• PyTorch (optional, for deep learning models)
• Astropy

Getting Started

# Clone the repository
git clone https://github.com/Sakeeb91/exoplanet-detection-ml.git
cd exoplanet-detection-ml

# Create virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Download sample data
python src/data/download_kaggle_subset.py

License

MIT License

References

• NASA Exoplanet Archive: https://exoplanetarchive.ipac.caltech.edu
• Lightkurve Documentation: https://docs.lightkurve.org
• Kepler Mission: https://www.nasa.gov/mission_pages/kepler/main/index.html