The Classification of Satellite Galaxies Based on Convolution Neural Networks Machine Learning Algorithm

Human exploration in space science unleashed positive achievements of mankind in understanding universe, the whole universe comprises of galaxies, nebula and black holes need to be identifiable in order to extend future exploration and better understanding of space, to leverage the space technology human intelligence does need a automation support in understanding cosmological formations, evolution’s and galactic activities, basic understanding of cosmology by pictorial information is main source of knowledge gathering of macroscopic world, technical application of satellites able to capture high resolution of images of planets and galaxies among which Hubble Space Telescope, Kepler Space Telescope, and Spitzer Space Telescopes embedded various types of imaging mechanisms such as infrared, photometer in capturing astronomical objects while the most advanced imaging techniques been applied in space observation laboratories to classify, understand pattern of shape, motion and temporal characteristics of space objects. In this research the data-set provided by NASA and European Space Agency of Galaxy Zoo challenge exists in Kaggle platform for classification of galaxy images with as the day passes by, the telescopes above the earth capture increasingly images of faraway galaxies. As telescopes bigger in size proceed on to collecting those images, the datasets begin to blow up in the size. For more understanding how the different shapes of galaxies connect to the domain of physics which has created them, such images need to be segregated and categorized using a Convolution Neural Network.