Identifying Exoplanets with Machine Learning Methods: A Preliminary Study

The discovery of habitable exoplanets has long been a heated topic in astronomy. Traditional methods for exoplanet identification include the wobble method, direct imaging, gravitational microlensing, etc., which not only require considerable investment manpower, time, and money, but also are limited by performance astronomical telescopes. In this study, we proposed idea using machine learning to identify exoplanets. We used Kepler dataset collected NASA from Space Observatory conduct supervised learning, predicts existence candidates as three-categorical classification task, decision tree, random forest, na\"ive Bayes, neural network; another consisted confirmed data unsupervised divides into different clusters, k-means clustering. As result, our models achieved accuracies 99.06%, 92.11%, 88.50%, 99.79%, respectively, task successfully obtained reasonable clusters task.