Learning Spatiotemporal Dynamical Systems from Point Process Observations
This addresses the limitation of current neural network-based approaches for modeling spatiotemporal systems under realistic, unconstrained observation conditions, which is incremental but important for applications like earthquake detection or pollution monitoring.
The paper tackles the problem of learning spatiotemporal dynamics from randomly collected point process observations, such as from sensor networks, by developing a new method that integrates neural differential equations, neural point processes, implicit neural representations, and amortized variational inference. The result is a model that outperforms existing methods with substantial improvements in predictive accuracy and computational efficiency on challenging datasets.
Spatiotemporal dynamics models are fundamental for various domains, from heat propagation in materials to oceanic and atmospheric flows. However, currently available neural network-based spatiotemporal modeling approaches fall short when faced with data that is collected randomly over time and space, as is often the case with sensor networks in real-world applications like crowdsourced earthquake detection or pollution monitoring. In response, we developed a new method that can effectively learn spatiotemporal dynamics from such point process observations. Our model integrates techniques from neural differential equations, neural point processes, implicit neural representations and amortized variational inference to model both the dynamics of the system and the probabilistic locations and timings of observations. It outperforms existing methods on challenging spatiotemporal datasets by offering substantial improvements in predictive accuracy and computational efficiency, making it a useful tool for modeling and understanding complex dynamical systems observed under realistic, unconstrained conditions.