Apollo XXI - an Astronomy Portal

Abstract

Astronomers have to deal with complex data to gain important insights which is a time-consuming task. Machine Learning techniques can help astronomers to analyze astronomical data in a simplified manner. The proposed web application consists of three different models. The first model can predict whether some meteor shower can be seen from a particular location along with the date and the name of the meteors. The proposed model is found to be 100% reliable in the experiments carried out so far. The second model can predict whether a celestial body is a candidate, confirmed or false positive instance of an exoplanet, based on the Kepler telescope data. This model uses random forest classifier and the accuracy achieved is 90.1%. The third model can predict whether there will be a delay in rocket launch according to different weather conditions. This model is based on decision tree classifier and has an accuracy of 98.3%.