在场层级对重子效应进行稳健边缘化以用于宇宙学推断
宇宙学与河外天体物理
2021-09-23 v1 星系天体物理
天体物理仪器与方法
计算机视觉与模式识别
机器学习
摘要
我们训练神经网络，从 CAMELS 项目数千个流体动力学模拟中包含总质量面密度的 二维图中执行无似然推断。我们表明，这些网络能从所有已分辨尺度（）提取超出一点函数与功率谱的信息，同时在场层级对重子物理进行稳健边缘化：该模型能从与训练所用完全不同的模拟中推断出 与 的值。
引用
@article{arxiv.2109.10360,
title = {Robust marginalization of baryonic effects for cosmological inference at the field level},
author = {Francisco Villaescusa-Navarro and Shy Genel and Daniel Angles-Alcazar and David N. Spergel and Yin Li and Benjamin Wandelt and Leander Thiele and Andrina Nicola and Jose Manuel Zorrilla Matilla and Helen Shao and Sultan Hassan and Desika Narayanan and Romeel Dave and Mark Vogelsberger},
journal= {arXiv preprint arXiv:2109.10360},
year = {2021}
}
备注
7 pages, 4 figures. Second paper of a series of four. The 2D maps, codes, and network weights used in this paper are publicly available at https://camels-multifield-dataset.readthedocs.io