Bayesian Recurrent Neural Network Models for Forecasting and Quantifying Uncertainty in Spatial-Temporal Data.
Bayesian Recurrent Neural Network Models for Forecasting and Quantifying Uncertainty in Spatial-Temporal Data.
复制标题
DOI：
10.3390/e21020184
复制
发表时间：
2019-02-15
期刊：
影响因子：
--
通讯作者：
Wikle CK
中科院分区：
文献类型：
--
作者：
McDermott PL;Wikle CK
递归神经网络（RNN）是机器学习和动态系统文献中常用的非线性动态模型，用于表示变量之间的复杂动态或顺序关系。最近，随着深度学习模型变得越来越普遍，RNN已被用于预测越来越复杂的系统。动态时空过程代表了一类复杂的系统，可以从这些类型的模型中受益。尽管RNN文献广泛且高度发达，但不确定性量化往往被忽视。即使在考虑时，不确定性通常也没有使用严格的框架（例如完全贝叶斯设置）进行量化。在这里，我们试图在一个更正式的框架中量化不确定性，同时保持预测的准确性，使这些模型具有吸引力，通过提出一个贝叶斯RNN模型的非线性时空预测。此外，我们对基本RNN进行了简单的修改，以帮助适应非线性时空数据的独特性质。该模型被应用到洛伦兹模拟和两个现实世界的非线性时空预测应用。
Recurrent neural networks (RNNs) are nonlinear dynamical models commonly used in the machine learning and dynamical systems literature to represent complex dynamical or sequential relationships between variables. Recently, as deep learning models have become more common, RNNs have been used to forecast increasingly complicated systems. Dynamical spatio-temporal processes represent a class of complex systems that can potentially benefit from these types of models. Although the RNN literature is expansive and highly developed, uncertainty quantification is often ignored. Even when considered, the uncertainty is generally quantified without the use of a rigorous framework, such as a fully Bayesian setting. Here we attempt to quantify uncertainty in a more formal framework while maintaining the forecast accuracy that makes these models appealing, by presenting a Bayesian RNN model for nonlinear spatio-temporal forecasting. Additionally, we make simple modifications to the basic RNN to help accommodate the unique nature of nonlinear spatio-temporal data. The proposed model is applied to a Lorenz simulation and two real-world nonlinear spatio-temporal forecasting applications.
登录
查看更多内容
影响因子：
6
作者：
Chandra, Rohitash;Zhang, Mengjie
通讯作者：
Zhang, Mengjie
影响因子：
1.7
作者：
Gladish, D. W.;Wikle, C. K.
通讯作者：
Wikle, C. K.
影响因子：
4.5
作者：
Hobert, James P.;Marchev, Dobrin
通讯作者：
Marchev, Dobrin
影响因子：
1.4
作者：
Dixon, Matthew F.;Polson, Nicholas G.;Sokolov, Vadim O.
通讯作者：
Sokolov, Vadim O.
DOI：
10.1073/pnas.1512080112
发表时间：
2015-08-11
影响因子：
11.1
作者：
Chorin, Alexandre J.;Lu, Fei
通讯作者：
Lu, Fei