HawkesNest: A Multi-Axis Synthetic Benchmark for Spatiotemporal Pattern Complexity
2026-06-15 • Machine Learning
Machine Learning
AI summaryⓘ
The authors created HawkesNest, a new way to test spatiotemporal point process (STPP) models using data with known complexity. They designed four difficulty factors related to space, time, and interactions, which can be adjusted to see how models handle different challenges. By doing this, the authors can better understand when and why models like Hawkes-family and neural methods fail. HawkesNest helps reveal specific weaknesses in models that are otherwise hard to detect with real-world data.
spatiotemporal point processHawkes processmodel evaluationgenerative modelcomplexity axesbackground heterogeneitycross-type interactionneural modelsspace-time entanglement
Authors
Yahya Aalaila, Sumantrak Mukherjee, Gerrit Großmann, Sebastian Vollmer
Abstract
Evaluation of spatiotemporal point process (STPP) models relies heavily on opaque real-world datasets, where latent generative structure is unknown and model failures are difficult to attribute. We introduce HawkesNest, a generator-aligned benchmark for controlled spatiotemporal pattern complexity built on a multivariate Hawkes backbone. HawkesNest defines four complexity axes: space--time entanglement, background heterogeneity, cross-type interaction, and domain topology. Each axis is associated with a deterministic index computed from the latent data-generating mechanism. By varying these axes while holding global rate, stability, and simulation budget fixed, HawkesNest enables diagnostic stress tests of STPP models under known structural difficulty. We verify that the indices are monotone and nearly orthogonal under controlled sweeps. We illustrate its use by showing that Hawkes-family baselines degrade under joint heterogeneity--entanglement complexity, even though they are structurally aligned with the Hawkes data-generating backbone. We further show that HawkesNest exposes neural-model sensitivity: AutoSTPP remains vulnerable under isolated increases in space--time entanglement. Code. Available at https://github.com/YahyaAalaila/HawkesNest