LINNA：Likestive推理神经网络加速器
LINNA: Likelihood Inference Neural Network Accelerator
摘要：
现代多探针宇宙学分析的贝叶斯后验推理会产生巨大的计算成本。例如，根据探测器的组合，暗能量调查（DES）数据的单一后验推断具有1到21天的挂钟时间，使用具有100个核心的最先进的计算集群。这些计算成本会对环境产生严重影响，而且长时间的挂钟会降低科学生产力。为了解决这些困难，我们引入了LINNA：Likestive推理神经网络加速器。相对于基线DES分析，LINNA降低了计算成本与后验推理的一个因素的8- 50。如果应用于鲁宾天文台遗产时空巡天（LSST Y1）的第一年宇宙学分析，我们保守估计LINNA将比美国$\$300,000$节省更多的能源成本，同时减少$\rm{CO}_2$排放$2,400$吨。为了实现这些约简，LINNA自动构建训练数据集，创建神经网络代理模型，并生成对后验进行采样的马尔可夫链。我们明确验证，LINNA准确地再现了第一年DES（DES Y1）宇宙学约束来自各种不同的数据向量与我们的默认代码设置，而不需要重新调整算法每次。此外，我们发现LINNA足以使LSST Y10多探针分析的准确和有效的采样。我们在https：//github.com/chto/linna上公开了LINNA，以使其他人能够在当代宇宙学分析中进行快速准确的后验推理。
英文摘要：
Bayesian posterior inference of modern multi-probe cosmological analyses incurs massive computational costs. For instance, depending on the combinations of probes, a single posterior inference for the Dark Energy Survey (DES) data had a wall-clock time that ranged from 1 to 21 days using a state-of-the-art computing cluster with 100 cores. These computational costs have severe environmental impacts and the long wall-clock time slows scientific productivity. To address these difficulties, we introduce LINNA: the Likelihood Inference Neural Network Accelerator. Relative to the baseline DES analyses, LINNA reduces the computational cost associated with posterior inference by a factor of 8--50. If applied to the first-year cosmological analysis of Rubin Observatory's Legacy Survey of Space and Time (LSST Y1), we conservatively estimate that LINNA will save more than US $\$300,000$ on energy costs, while simultaneously reducing $\rm{CO}_2$ emission by $2,400$ tons. To accomplish these reductions, LINNA automatically builds training data sets, creates neural network surrogate models, and produces a Markov chain that samples the posterior. We explicitly verify that LINNA accurately reproduces the first-year DES (DES Y1) cosmological constraints derived from a variety of different data vectors with our default code settings, without needing to retune the algorithm every time. Further, we find that LINNA is sufficient for enabling accurate and efficient sampling for LSST Y10 multi-probe analyses. We make LINNA publicly available at https://github.com/chto/linna, to enable others to perform fast and accurate posterior inference in contemporary cosmological analyses.
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。
