A volumetric deep Convolutional Neural Network for simulation of mock dark matter halo catalogues

Abstract

For modern large-scale structure survey techniques it has become standard practice to test data analysis pipelines on large suites of mock simulations, a task which is currently prohibitively expensive for full N-body simulations. Instead of calculating this costly gravitational evolution, we have trained a three-dimensional deep Convolutional Neural Network (CNN) to identify dark matter protohaloes directly from the cosmological initial conditions. Training on halo catalogues from the Peak Patch semi-analytic code, we test various CNN architectures and find they generically achieve a Dice coefficient of ∼92 per cent in only 24 h of training. We present a simple and fast geometric halo finding algorithm to extract haloes from this powerful pixel-wise binary classifier and find that the predicted catalogues match the mass function and power spectra of the ground truth simulations to within ∼10 per cent. We investigate the effect of long-range tidal forces on an object-by-object basis and find that the network’s predictions are consistent with the non-linear ellipsoidal collapse equations used explicitly by the Peak Patch algorithm.