arXiv:2410.07548stat.MLastro-ph.CO2024-10中稿 · NeurIPS被引 9

用神经网络增强传统统计量,提升低数据量下的物理推断精度

Hybrid Summary Statistics

  • 结合领域知识与神经网络输出,最大化互信息提取信息
  • 在低数据量下显著优于纯神经网络或简单拼接方法
  • 适用于宇宙学中非高斯参数的高精度推断

我们提出一种方法,从参数空间稀疏采样的训练集中捕捉高信息量的后验分布,以实现鲁棒的模拟驱动推断。在物理推断问题中,可利用领域知识定义传统摘要统计量来捕获数据的部分信息。我们证明,通过将这些统计量与神经网络输出相结合,以最大化互信息,相比仅使用神经摘要或与现有统计量拼接,能显著提升信息提取能力,并在训练数据稀缺时保持推断稳健性。本文引入了两种损失形式实现该目标,并将其应用于两个不同的宇宙学数据集,成功提取了非高斯参数信息。

原文摘要 · Abstract (English)

We present a way to capture high-information posteriors from training sets that are sparsely sampled over the parameter space for robust simulation-based inference. In physical inference problems, we can often apply domain knowledge to define traditional summary statistics to capture some of the information in a dataset. We show that augmenting these statistics with neural network outputs to maximise the mutual information improves information extraction compared to neural summaries alone or their concatenation to existing summaries and makes inference robust in settings with low training data. We introduce 1) two loss formalisms to achieve this and 2) apply the technique to two different cosmological datasets to extract non-Gaussian parameter information.

推断神经网络宇宙学摘要统计

Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品;中文卡片由大模型生成,请以原文为准。