Automatic Differentiation for Gradient Estimators in Simulation
Automatic Differentiation for Gradient Estimators in Simulation
复制标题
DOI：
10.1109/wsc57314.2022.10015421
复制
发表时间：
2022-12
期刊：
影响因子：
--
通讯作者：
Matthew T. Ford;S. Henderson;David J. Eckman
中科院分区：
文献类型：
--
作者：
Matthew T. Ford;S. Henderson;David J. Eckman
自动微分(AD)可以直接从仿真代码提供无穷小扰动分析(IPA)导数估计。这些梯度估计器很容易通过解析获得，至少在原则上是这样，但在代码中推导和实现可能会很繁琐。AD软件工具旨在通过编写模拟代码来减轻这种工作量。我们回顾了选择AD工具进行仿真时的注意事项，演示了如何将一些特定的AD工具应用于仿真，并提供了有洞察力的实验，突出了在仿真中应用AD时所做的不同选择的效果。
Automatic differentiation (AD) can provide infinitesimal perturbation analysis (IPA) derivative estimates directly from simulation code. These gradient estimators are simple to obtain analytically, at least in principle, but may be tedious to derive and implement in code. AD software tools aim to ease this workload by requiring little more than writing the simulation code. We review considerations when choosing an AD tool for simulation, demonstrate how to apply some specific AD tools to simulation, and provide insightful experiments highlighting the effects of different choices to be made when applying AD in simulation.