Neural Spatio-Temporal Point Processes
This work addresses the need for accurate event modeling in fields like urban mobility and neuroscience, representing an incremental improvement through novel neural architectures.
The authors tackled the problem of modeling discrete events in continuous time and space by proposing a new class of parameterizations for spatio-temporal point processes, achieving flexible, high-fidelity models validated across diverse datasets like seismology and epidemiology.
We propose a new class of parameterizations for spatio-temporal point processes which leverage Neural ODEs as a computational method and enable flexible, high-fidelity models of discrete events that are localized in continuous time and space. Central to our approach is a combination of continuous-time neural networks with two novel neural architectures, i.e., Jump and Attentive Continuous-time Normalizing Flows. This approach allows us to learn complex distributions for both the spatial and temporal domain and to condition non-trivially on the observed event history. We validate our models on data sets from a wide variety of contexts such as seismology, epidemiology, urban mobility, and neuroscience.