In the vast cosmic ocean of over 5,000 confirmed exoplanets, identifying which worlds might harbor life remains one of astronomy’s greatest challenges. The Habitable Planet Hunter tackles this needle-in-a-haystack problem by leveraging the comprehensive PHL Exoplanet Catalog to build machine learning models that can distinguish potentially habitable worlds from barren rocks and gas giants.
This Jupyter-based toolkit employs a carefully curated feature set of 30 stellar and planetary parameters—from orbital eccentricity and stellar luminosity to planetary surface gravity and host star metallicity. The system deliberately excludes pre-calculated habitability proxies like Earth Similarity Index, forcing models to learn the fundamental physical relationships that govern habitability: how stellar radiation interacts with orbital distance, how planetary mass affects atmospheric retention, and how stellar age influences the evolution of planetary atmospheres. The framework supports both binary and multi-class classification, allowing researchers to explore conservative versus optimistic habitability scenarios.
Built for both seasoned astrobiologists and machine learning practitioners venturing into space science, this tool democratizes habitability research beyond major observatories. Whether you’re validating theoretical models against observational data or exploring feature importance in exoplanet characterization, the Habitable Planet Hunter provides a robust foundation for understanding what makes a world potentially suitable for life as we know it.
⭐ Stars: 6
💻 Language: Jupyter Notebook
🔗 Repository: opencodeiiita/habitable-planet-hunter