The investigation of exoplanetary habitability is integral to advancing our knowledge of extraterrestrial life potential and detailing the environmental conditions of distant worlds. In this analysis, we explore the properties of exoplanets situated with respect to circumstellar habitable zones by implementing a sophisticated filtering methodology on data from the NASA Exoplanet Archive. This research encompasses a thorough examination of 5595 confirmed exoplanets listed in the Archive as of 10 March 2024, systematically evaluated according to their calculated average surface temperatures and stellar classifications of their host stars, taking into account the biases implicit in the methodologies used for their discovery. Machine learning, in the form of a Random Forest classifier and an XGBoost classifier, is applied in the classification with high accuracies. The feature importance analysis indicates that our approach captures the most important parameters for habitability classification. Our findings elucidate distinctive patterns in exoplanetary attributes, which are significantly shaped by the spectral classifications and mass of the host stars. The insights garnered from our study both inform refinement of existing models for managing burgeoning exoplanetary datasets, and lay foundational groundwork for more in-depth explorations of the dynamic relationships between exoplanets and their stellar environments.
Read full abstract