Classification of gravitational waves from supernova core collapses using affinity propagation

Numerical relativistic simulations of a supernova core collapse indicate that a gravitational waveform from the supernova core collapse has several types. The gravitational waveform is strongly related to the stellar structure such as a rotation of the supernova core and an equation of state. Knowing the distribution of the types of the gravitational waveforms is an probe to elucidate the story of evolution of the supernova. In this paper, we propose a clustering method to determine the distribution. We performed Monte Carlo simulations of classification of simulated gravitational waves from supernovae using the affinity propagation and showed the efficiency of the method. We found that the classification by the affinity propagation worked well if a supernova core collapse occurs in our Galaxy.