Deep Random Features for Scalable Interpolation of Spatiotemporal Data
This addresses the need for scalable and expressive models in earth observation systems, though it is incremental as it builds on prior work in Gaussian processes and deep learning.
The authors tackled the problem of scalable interpolation for spatiotemporal remote-sensing data by proposing a Bayesian deep learning approach using deep neural networks with random feature expansions, achieving competitive or superior results with well-calibrated uncertainties compared to existing methods.
The rapid growth of earth observation systems calls for a scalable approach to interpolate remote-sensing observations. These methods in principle, should acquire more information about the observed field as data grows. Gaussian processes (GPs) are candidate model choices for interpolation. However, due to their poor scalability, they usually rely on inducing points for inference, which restricts their expressivity. Moreover, commonly imposed assumptions such as stationarity prevents them from capturing complex patterns in the data. While deep GPs can overcome this issue, training and making inference with them are difficult, again requiring crude approximations via inducing points. In this work, we instead approach the problem through Bayesian deep learning, where spatiotemporal fields are represented by deep neural networks, whose layers share the inductive bias of stationary GPs on the plane/sphere via random feature expansions. This allows one to (1) capture high frequency patterns in the data, and (2) use mini-batched gradient descent for large scale training. We experiment on various remote sensing data at local/global scales, showing that our approach produce competitive or superior results to existing methods, with well-calibrated uncertainties.