Habitability of Exoplanets using Deep Learning

In observational astronomy, researchers have reached a point where large datasets from sky surveys are regularly published. Sending probes to land on asteroids, collect material, and ship it back to Earth is one strategy for space exploration. Astronomical data is rapidly increasing in size and complexity as space and ground-based telescopes are developed and deployed. Machine learning has gained popularity among astronomers in recent years, and it is now used to solve a variety of tasks, including classification, regression, clustering, outlier detection, time series analysis, association law, and so on. The study on previous work of papers on exoplanets and their habitability show machine learning methods including Support Vector Classification, Random forest, K-Nearest Neighbor, and so on were used in the majority of the papers. Until now, only a few deep learning methods have been studied. As a result, this research work overcomes different challenges faced by astronomers dealing with large data and seeking relevant knowledge for each objective using Deep Learning techniques. Furthermore, deep learning techniques are capable of managing complex data with ease. To begin, this paper proposes ASTRONET, a deep learning architecture, to look for exoplanets that are habitable based on their planet gravity, eccentricity, mass, radius, and other characteristics. This research paper explains the entire process of finding exoplanets and categorizing them based on their habitability. Finally, to establish a knowledge base of parameters that affect the habitability of exoplanets.