Large-scale Gravitational Lens Modeling with Bayesian Neural Networks for Accurate and Precise Inference of the Hubble Constant
Large-scale Gravitational Lens Modeling with Bayesian Neural Networks for Accurate and Precise Inference of the Hubble Constant
复制标题
DOI：
10.3847/1538-4357/abdfc4
复制
发表时间：
2020-11
期刊：
影响因子：
--
通讯作者：
Ji Won Park;S. Wagner-Carena;S. Birrer;P. Marshall;J. Lin;A. Roodman
中科院分区：
文献类型：
--
作者：
Ji Won Park;S. Wagner-Carena;S. Birrer;P. Marshall;J. Lin;A. Roodman
我们研究了用近似贝叶斯神经网络(BNN)对数百个时间延迟引力透镜进行建模以确定哈勃常数(H0)的方法。我们的BNN是在合成哈勃太空望远镜质量的图像上训练的，这些图像包括透镜星系光在内的强透镜活动星系核。该神经网络可以准确地刻画外剪切场中控制椭圆幂函数质量分布的模型参数的后验概率密度函数(PDF)。然后，我们使用来自可信的专用监测活动的模拟时间延迟测量，将BNN推断的后验PDF传播到集合H0推断中。假设测量好的时间延迟和合理的镜头环境先验设置，我们在推断的H0中实现了每个镜头9.3%的中位精度。一组200个测试镜头的简单组合得到了0.5kM的S−1Mpc−1(0.7%)的精度，在这次H0恢复测试中没有可检测到的偏差。对于200个镜头，整个流水线的计算时间--包括训练集的生成、BNN训练和H0推理--转化为每个镜头平均9分钟，随着样本大小的增加，收敛到每个镜头6分钟。由于完全自动化和高效，我们的流水线是在H0推断的透镜建模中探索系综级系统学的一个很有前途的工具。
We investigate the use of approximate Bayesian neural networks (BNNs) in modeling hundreds of time delay gravitational lenses for Hubble constant (H 0) determination. Our BNN was trained on synthetic Hubble Space Telescope quality images of strongly lensed active galactic nuclei with lens galaxy light included. The BNN can accurately characterize the posterior probability density functions (PDFs) of model parameters governing the elliptical power-law mass profile in an external shear field. We then propagate the BNN-inferred posterior PDFs into an ensemble H 0 inference, using simulated time delay measurements from a plausible dedicated monitoring campaign. Assuming well-measured time delays and a reasonable set of priors on the environment of the lens, we achieve a median precision of 9.3% per lens in the inferred H 0. A simple combination of a set of 200 test lenses results in a precision of 0.5 km s−1 Mpc−1 (0.7%), with no detectable bias in this H 0 recovery test. The computation time for the entire pipeline—including the generation of the training set, BNN training and H 0 inference—translates to 9 minutes per lens on average for 200 lenses and converges to 6 minutes per lens as the sample size is increased. Being fully automated and efficient, our pipeline is a promising tool for exploring ensemble-level systematics in lens modeling for H 0 inference.