STACI: Spatio-Temporal Aleatoric Conformal Inference
This addresses the need for interpretable and statistically valid uncertainty quantification in spatio-temporal modeling, with incremental improvements in scalability and accuracy over Gaussian Processes and deep learning methods.
The authors tackled the problem of scalable and accurate uncertainty quantification for spatio-temporal fields by proposing STACI, a framework combining variational Bayesian neural networks with conformal inference, which outperformed existing methods and scaled to millions of observations.
Fitting Gaussian Processes (GPs) provides interpretable aleatoric uncertainty quantification for estimation of spatio-temporal fields. Spatio-temporal deep learning models, while scalable, typically assume a simplistic independent covariance matrix for the response, failing to capture the underlying correlation structure. However, spatio-temporal GPs suffer from issues of scalability and various forms of approximation bias resulting from restrictive assumptions of the covariance kernel function. We propose STACI, a novel framework consisting of a variational Bayesian neural network approximation of non-stationary spatio-temporal GP along with a novel spatio-temporal conformal inference algorithm. STACI is highly scalable, taking advantage of GPU training capabilities for neural network models, and provides statistically valid prediction intervals for uncertainty quantification. STACI outperforms competing GPs and deep methods in accurately approximating spatio-temporal processes and we show it easily scales to datasets with millions of observations.