Deep Learning for Cosmological Parameter Inference from Dark Matter Halo Density Field

We propose a lightweight deep convolutional neural network (lCNN) to estimate cosmological parameters from simulated three-dimensional DM halo distributions and associated statistics. The training dataset comprises 2000 realizations of a cubic box with a side length of 1000 h⁻¹ Mpc, and interpolated over a cubic grid of 300³ voxels, with each simulation produced using 512³ DM particles and 512³ neutrinos . Under the flat ΛCDM model, simulations vary standard six cosmological parameters including Ωₘ, Ωb, h, nₛ, σ₈, w, along with the neutrino mass sum, M_ν. We find that: 1) within the framework of lCNN, extracting large-scale structure information is more efficient from the halo density field compared to relying on the statistical quantities including the power spectrum, the two-point correlation function, and the coefficients from wavelet scattering transform; 2) combining the halo density field with its Fourier transformed counterpart enhances predictions, while augmenting the training dataset with measured statistics further improves performance; 3) achieving high accuracy in inferring Ωₘ, h, nₛ, and σ₈ by the neural network model, while being inefficient in predicting Ωb,M_ν and w; 4) compared to the simple random forest network trained with three statistical quantities, lCNN yields unbiased estimations with reduced statistical errors: approximately 33.3% for Ωₘ, 20.0% for h, 8.3% for nₛ, and 40.0% for σ₈. Our study emphasizes this lCNN-based novel approach in extracting large-scale structure information and estimating cosmological parameters.