Scalable Spatiotemporal Prediction with Bayesian Neural Fields
This addresses the need for scalable statistical methods in applications like air pollution monitoring and disease tracking, though it appears incremental as it builds on existing neural and Bayesian approaches.
The paper tackles the challenge of modeling large-scale spatiotemporal data by introducing the Bayesian Neural Field (BayesNF), which combines deep neural networks with Bayesian inference for flexible and scalable predictions, showing improvements on climate and public health datasets with tens to hundreds of thousands of measurements.
Spatiotemporal datasets, which consist of spatially-referenced time series, are ubiquitous in diverse applications, such as air pollution monitoring, disease tracking, and cloud-demand forecasting. As the scale of modern datasets increases, there is a growing need for statistical methods that are flexible enough to capture complex spatiotemporal dynamics and scalable enough to handle many observations. This article introduces the Bayesian Neural Field (BayesNF), a domain-general statistical model that infers rich spatiotemporal probability distributions for data-analysis tasks including forecasting, interpolation, and variography. BayesNF integrates a deep neural network architecture for high-capacity function estimation with hierarchical Bayesian inference for robust predictive uncertainty quantification. Evaluations against prominent baselines show that BayesNF delivers improvements on prediction problems from climate and public health data containing tens to hundreds of thousands of measurements. Accompanying the paper is an open-source software package (https://github.com/google/bayesnf) that runs on GPU and TPU accelerators through the JAX machine learning platform.