Statistical and Machine Learning Perspectives on Exoplanets Prediction and Classification

Exoplanets are the planets orbiting stars out of our solar system. The exploration of such planets has become a growing field in astronomy but the analysis of the data related to such planets involves many challenges when traditional methods are used as this data is often vast and is obtained with the help of telescopes and space missions. This study proposes a methodology that aims to classify exoplanets, taken from datasets from the Kepler Objects of Interest (KOI) database and the NASA Exoplanet archive, as confirmed, candidate, or false positives. This approach makes use of modern analytical techniques, such as statistical analysis, probability theory, hypothesis testing, and machine learning (ML) algorithms, to facilitate the detection and classification of exoplanets. Clustering methods such as BIRCH and K-Means are also performed on the Planetary data. This study contributes to improving our knowledge and understanding of exoplanetary systems, thereby highlighting the relevance of modern analytical tools in the field of astronomy.