Asymptotically unbiased estimation of physical observables with neural samplers
Asymptotically unbiased estimation of physical observables with neural samplers
复制标题
DOI：
10.1103/physreve.101.023304
复制
发表时间：
2020-02-10
影响因子：
2.4
通讯作者：
Kessel, Pan
中科院分区：
文献类型：
--
作者：
Nicoli, Kim A.;Nakajima, Shinichi;Kessel, Pan
我们提出了一个使用生成性神经采样器估计可观测值的一般框架，着重于提供精确采样概率的现代深度生成性神经网络。在这个框架下，我们给出了一般可观量的渐近无偏估计，包括那些显式依赖于配分函数的估计，如自由能或熵，并得到了相应的方差估计。通过对二维伊辛模型的数值实验，验证了该方法的实用性，突出了其相对于现有方法的优越性。我们的方法极大地提高了产生式神经采样器对真实世界物理系统的适用性。
We propose a general framework for the estimation of observables with generative neural samplers focusing on modern deep generative neural networks that provide an exact sampling probability. In this framework, we present asymptotically unbiased estimators for generic observables, including those that explicitly depend on the partition function such as free energy or entropy, and derive corresponding variance estimators. We demonstrate their practical applicability by numerical experiments for the two-dimensional Ising model which highlight the superiority over existing methods. Our approach greatly enhances the applicability of generative neural samplers to real-world physical systems.