Cosmological parameter estimation from large-scale structure deep learning
- PDF / 5,737,054 Bytes
- 15 Pages / 595.28 x 793.7 pts Page_size
- 13 Downloads / 235 Views
. Article .
November 2020 Vol. 63 No. 11: 110412 https://doi.org/10.1007/s11433-020-1586-3
Cosmological parameter estimation from large-scale structure deep learning ShuYang Pan1, MiaoXin Liu1, Jaime Forero-Romero2, Cristiano G. Sabiu3, ZhiGang Li4, HaiTao Miao1, and Xiao-Dong Li1* 1 School
of Physics and Astronomy, Sun Yat-Sen University, Guangzhou 510297, China; de F´ısica, Universidad de los Andes, Bogot´a CP 111711, Colombia; 3 Department of Astronomy, Yonsei University, Seoul 03722, Korea; 4 College of Physics and Electronic Engineering, Nanyang Normal University, Nanyang 473061, China 2 Departamento
Received March 16, 2020; accepted June 1, 2020; published online September 11, 2020
We propose a light-weight deep convolutional neural network (CNN) to estimate the cosmological parameters from simulated 3-dimensional dark matter distributions with high accuracy. The training set is based on 465 realizations of a cubic box with a side length of 256 h−1 Mpc, sampled with 1283 particles interpolated over a cubic grid of 1283 voxels. These volumes have cosmological parameters varying within the flat ΛCDM parameter space of 0.16 ≤ Ωm ≤ 0.46 and 2.0 ≤ 109 As ≤ 2.3. The neural network takes as an input cubes with 323 voxels and has three convolution layers, three dense layers, together with some batch normalization and pooling layers. In the final predictions from the network we find a 2.5% bias on the primordial amplitude σ8 that cannot easily be resolved by continued training. We correct this bias to obtain unprecedented accuracy in the cosmological parameter estimation with statistical uncertainties of δΩm =0.0015 and δσ8 =0.0029, which are several times better than the results of previous CNN works. Compared with a 2-point analysis method using the clustering region of 0-130 and 10-130 h−1 Mpc, the CNN constraints are several times and an order of magnitude more precise, respectively. Finally, we conduct preliminary checks of the error-tolerance abilities of the neural network, and find that it exhibits robustness against smoothing, masking, random noise, global variation, rotation, reflection, and simulation resolution. Those effects are well understood in typical clustering analysis, but had not been tested before for the CNN approach. Our work shows that CNN can be more promising than people expected in deriving tight cosmological constraints from the cosmic large scale structure. large-scale structure of Universe, cosmological parameters, dark energy, machine learning PACS number(s): Citation:
98.80.-k, 98.80.Es, 95.36.+x
S. Y. Pan, M. X. Liu, J. Forero-Romero, C. G. Sabiu, Z. G. Li, H. T. Miao, and X.-D. Li, Cosmological parameter estimation from large-scale structure deep learning, Sci. China-Phys. Mech. Astron. 63, 110412 (2020), https://doi.org/10.1007/s11433-020-1586-3
1 Introduction The current standard model of cosmology has been highly successful at describing the Universe on large scales. From the anisotropic temperature fluctuations in the cosmic microwave background (CMB) to the late time clus
Data Loading...
