arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2206.01340hep-phnucl-th
收藏

基于信息场的喷注输运系数全局贝叶斯推断

Information field based global Bayesian inference of the jet transport coefficient

Man Xie, Weiyao Ke, Hanzhong Zhang, Xin-Nian Wang

AI总结:

本文提出无长程关联的信息场方法,首次利用RHIC和LHC全部相关实验数据对喷注输运系数随温度的变化进行全局贝叶斯推断,揭示了其强烈的温度依赖性并避免了函数形式偏置。

摘要:

贝叶斯统计推断是进行模型-数据比较以及提取物理参数的有力工具,而这些物理参数往往是系统变量的未知函数。现有的贝叶斯分析通常依赖于对未知函数的显式参数化。这种做法可能引入长程关联,从而在实验数据未探测的变量空间区域对物理参数施加人为的约束。我们发展了一种信息场(IF)方法来建模未知函数的先验分布,该方法不受长程关联的影响。我们将该IF方法应用于喷注输运系数$\hat q$作为温度($T$)的函数的首次全局贝叶斯推断,所用数据包括RHIC和LHC能区重离子碰撞中现有的全部单举强子、双强子以及$γ$-强子谱实验数据。提取得到的$\hat q/T^3$表现出强烈的$T$依赖性,这一结果来源于当逐步纳入更中心碰撞和更高碰撞能量的数据时,约束能力不断增强。IF方法保证了所提取的$T$依赖性不受特定函数形式的偏置影响。

英文摘要:

Bayesian statistical inference is a powerful tool for model-data comparisons and extractions of physical parameters that are often unknown functions of system variables. Existing Bayesian analyses often rely on explicit parametrizations of the unknown function. It can introduce long-range correlations that impose fictitious constraints on physical parameters in regions of the variable space that are not probed by the experimental data. We develop an information field (IF) approach to modeling the prior distribution of the unknown function that is free of long-range correlations. We apply the IF approach to the first global Bayesian inference of the jet transport coefficient $\hat q$ as a function of temperature ($T$) from all existing experimental data on single-inclusive hadron, di-hadron and $γ$-hadron spectra in heavy-ion collisions at RHIC and LHC energies. The extracted $\hat q/T^3$ exhibits a strong $T$-dependence as a result of the progressive constraining power when data from more central collisions and at higher colliding energies are incrementally included. The IF method guarantees that the extracted $T$-dependence is not biased by a specific functional form.

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。

↑