当前位置: X-MOL 学术arXiv.cs.NA › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Mutual Information for Explainable Deep Learning of Multiscale Systems
arXiv - CS - Numerical Analysis Pub Date : 2020-09-07 , DOI: arxiv-2009.04570
S{\o}ren Taverniers and Eric J. Hall and Markos A. Katsoulakis and Daniel M. Tartakovsky

Timely completion of design cycles for multiscale and multiphysics systems ranging from consumer electronics to hypersonic vehicles relies on rapid simulation-based prototyping. The latter typically involves high-dimensional spaces of possibly correlated control variables (CVs) and quantities of interest (QoIs) with non-Gaussian and/or multimodal distributions. We develop a model-agnostic, moment-independent global sensitivity analysis (GSA) that relies on differential mutual information to rank the effects of CVs on QoIs. Large amounts of data, which are necessary to rank CVs with confidence, are cheaply generated by a deep neural network (DNN) surrogate model of the underlying process. The DNN predictions are made explainable by the GSA so that the DNN can be deployed to close design loops. Our information-theoretic framework is compatible with a wide variety of black-box models. Its application to multiscale supercapacitor design demonstrates that the CV rankings facilitated by a domain-aware Graph-Informed Neural Network are better resolved than their counterparts obtained with a physics-based model for a fixed computational budget. Consequently, our information-theoretic GSA provides an "outer loop" for accelerated product design by identifying the most and least sensitive input directions and performing subsequent optimization over appropriately reduced parameter subspaces.

中文翻译:

多尺度系统可解释深度学习的互信息

及时完成从消费电子产品到高超音速飞行器的多尺度和多物理系统的设计周期依赖于基于快速仿真的原型设计。后者通常涉及可能相关的控制变量 (CV) 和具有非高斯和/或多模态分布的感兴趣量 (QoI) 的高维空间。我们开发了一种与模型无关、与矩无关的全局敏感性分析 (GSA),它依赖于差分互信息来对 CV 对 QoI 的影响进行排序。大量数据是自信地对 CV 进行排名所必需的,这些数据是由底层过程的深度神经网络 (DNN) 代理模型廉价生成的。GSA 可以解释 DNN 预测,以便可以部署 DNN 以关闭设计循环。我们的信息论框架与各种黑盒模型兼容。它在多尺度超级电容器设计中的应用表明,与使用基于物理的模型在固定计算预算下获得的对应物相比,由域感知图通知神经网络促进的 CV 排名得到了更好的解决。因此,我们的信息论 GSA 通过识别最敏感和最不敏感的输入方向并在适当减少的参数子空间上执行后续优化,为加速产品设计提供了一个“外循环”。它在多尺度超级电容器设计中的应用表明,与使用基于物理的模型在固定计算预算下获得的对应物相比,由域感知图通知神经网络促进的 CV 排名得到了更好的解决。因此,我们的信息论 GSA 通过识别最敏感和最不敏感的输入方向并在适当减少的参数子空间上执行后续优化,为加速产品设计提供了一个“外循环”。它在多尺度超级电容器设计中的应用表明,与使用基于物理的模型在固定计算预算下获得的对应物相比,由域感知图通知神经网络促进的 CV 排名得到了更好的解决。因此,我们的信息论 GSA 通过识别最敏感和最不敏感的输入方向并在适当减少的参数子空间上执行后续优化,为加速产品设计提供了一个“外循环”。
更新日期:2020-09-11
down
wechat
bug