Artificial neural network based clustering for analysis of astrophysical spectra

Abstract

One of the largest sources of big data is data from space observatories which are spectral, optical and numerical in nature. The analysis of this data is often performed using conventional computing techniques utilising vast computational resources, typically supercomputers. Brute force algorithms often play a part in such computations. In this paper, a novel artificial neural network based clustering approach is proposed for spectral classification of stellar data.