Collaboration Workshop on Cosmology and Galaxy Formation

Session

Simulation & inference

Jun 22, 2023, 9:00 AM

3rd floor meeting room (SJTU & Suzhou Bay)

Conveners

Simulation & inference: constrained simulation

• Hong Guo (Shanghai Astronomical Observatory)

Simulation & inference: Machine learning

• Yin Li

29. Reconstructing and stacking, a promising way to study galaxy formation and evolution

Prof. Houjun Mo (U. Mass.)

6/22/23, 9:00 AM

Inference

Talk

I will present a review on how to use targets provided by galaxies, galaxy groups and reconstructed density field to study the cosmic web, galaxy formation and evolution.

46. Connection between SDSS galaxies and ELUCID subhalos in the eye of machine learning

Xiaoju Xu (上海交通大学)

6/22/23, 9:25 AM

Halo

Talk

Machine learning techniques are widely implemented to predict galaxy properties based on halo or subhalo properties in semi-analytic models (SAM) and hydrodynamic simulations. The multivariate relation between galaxy and halo or subhalo can be captured and reproduced efficiently. However, the galaxy formation and evolution in galaxy formation models may deviate from those in the real Universe,...

33. Using Constrained Hydro Simulations to Explore the Gas Properties of the Cosmic Web

Renjie Li

6/22/23, 9:45 AM

Gas

Talk

Using reconstructed initial conditions in the Sloan Digital Sky Survey (SDSS) survey volume, we carry out constrained hydrodynamic simulations in three regions representing different types of the cosmic web: the Coma cluster of galaxies; the SDSS Great Wall; and a large low-density region at z ∼ 0.05. The simulations successfully predict some discontinuities associated with shock fronts and...

8. Differentiable Cosmological Simulation with Adjoint Method

Yin Li

6/22/23, 10:05 AM

Simulation

Talk

Rapid advances in deep learning have brought not only myriad powerful neural networks, but also breakthroughs that benefit established scientific research. In particular, automatic differentiation (AD) tools and computational accelerators like GPUs have facilitated forward modeling of the Universe with differentiable simulations. Current differentiable cosmological simulations are limited by...

12. Analyzing the large-scale structure of the Universe with the assistance of machine learning techniques

霄栋 李 (中山大学)

6/22/23, 10:45 AM

Inference

Talk

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...

50. DarkAI: Mapping the large-scale density field of dark matter using AI

Feng Shi (Xidian University)

6/22/23, 11:05 AM

Inference

Talk

We develop a deep learning technique to reconstruct the dark matter density field from the redshift-space distribution of dark matter halos. We implement a UNet-architecture neural network and successfully trained it using the COLA fast simulation, which is an approximation of the N-body simulation with $512^3$ particles in a box size of $500 \mpch$. We evaluate the resulting UNet model not...

47. Estimation of photo-z probability density functions via deep learning with statistical basis

Qiufan Lin (Pengcheng Lab)

6/22/23, 11:25 AM

Inference

Talk

Obtaining well-calibrated probability density functions (PDFs) of photometric redshift (photo-z) for galaxies without using spectroscopy remains a challenge for many science goals. Deep learning tools have proven to be powerful for this task and gained growing popularity. These include, in particular, state-of-the-art deep neural networks that are typically fed with multi-band galaxy images or...

56. Relating galaxies across different redshift to study galaxy evolution

Kai Wang

6/22/23, 11:45 AM