Speaker
Description
I will discuss our work on the reconstruction of the cosmic velocity field from the redshift-space distribution of dark matter halos using the state-of-the-art deep learning technique. We were able to accurately recover the magnitude, divergence and vorticity of the velocity field, with the power spectra recovered at 80% accuracy at $k < 1.1\ h$/Mpc. This approach is very promising and presents an alternative method to correct the redshift-space distortions using the measured 3D spatial information of halos. Additionally, I will present our recent work on the estimation of cosmological parameters using deep learning algorithms. Our work shows that machine learning technique has the potential to greatly improve our understanding of the Universe and its evolution.
