贝叶斯机器学习对宇宙不透明度的约束：$H_{0}$张力问题的隐藏面
Constraints on Cosmic Opacity from Bayesian Machine Learning: The hidden side of the $H_{0}$ tension problem
摘要：
贝叶斯（概率）机器学习用于探测宇宙的不透明性。它依赖于一个生成过程，其中模型是生成涉及模型未知参数、我们的先验信念的数据的关键对象，并允许我们获得后验结果。宇宙不透明度的约束条件是针对两个平坦模型$Λ$ CDM和XCDM（这两个模型都有$ω_{de} \neq -1$），分别针对三个红移范围$z\in[0,2.5]$、$z \in[0,5]$和$z\in[0,10]$确定的。这是为了了解在非常深的宇宙中，对宇宙不透明度的约束是如何变化的，也是为了检查红移范围的依赖程度。考虑以下不透明度形式，$τ(z) = 2εz$和$τ(z) = (1+z)^{2ε} -1$，对应于在$z=0$的观察者和在$z$的源。我们的分析结果表明，宇宙不是完全透明的，这可能对$H_{0}$张力问题产生重大影响。在生成过程中，由于宇宙的不透明性，观测者接收到的通量减少的事实已经被考虑在内。在分析中，光度距离与宇宙学模型已被采用。
英文摘要：
Bayesian (Probabilistic) Machine Learning is used to probe the opacity of the Universe. It relies on a generative process where the model is the key object to generate the data involving the unknown parameters of the model, our prior beliefs, and allows us to get the posterior results. The constraints on the cosmic opacity are determined for two flat models, $Λ$CDM and XCDM (this having $ω_{de} \neq -1$), for three redshift ranges, $z\in[0,2.5]$, $z \in[0,5]$, and $z\in[0,10]$, in each case. This is to understand how the constraints on the cosmic opacity could change in the very deep Universe, and also to check to what extent there is a redshift-range dependence. The following forms for the opacity, $τ(z) = 2εz$ and $τ(z) = (1+z)^{2ε} -1$, corresponding to an observer at $z=0$ and a source at $z$, are considered. The results of our analysis show that the Universe is not fully transparent, and this may have a significant impact on the $H_{0}$ tension problem. In the generative process, the fact that, owing to cosmic opacity, the flux received by the observer is reduced has been taken into account. In the analysis, the luminosity distance associated with the cosmological model has been employed.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
