YSPH Biostatistics Seminar: "Three Challenges Confronting Spatiotemporal Hawkes Models"

NOTE: BIS 525 students are required to attend in person (47 College St., Room 106A). All others are requested to attend via Zoom.

SPEAKER: Andrew Holbrook, PhD, Assistant Professor, Department of Biostatistics, University of California, Los Angeles

TITLE: "Three Challenges Confronting Spatiotemporal Hawkes Models"

ABSTRACT: Hawkes processes are useful for modeling self-exciting phenomena such as earthquakes, wildfires, gun violence, viral outbreaks and social network activity. Unsurprisingly, the stochastic process models are the subject of intense interest within the data science community. Dr Holbrook will share big picture lessons learned applying spatiotemporal Hawkes processes to the analysis of gunfire, wildfire and Ebola virus data. In particular, three barriers stand in the way of the application of Hawkes models within meaningful 21st century science: (1) big data, (2) spatial data precision and (3) big models.

Associated publications:

1. Holbrook, Andrew J., et al. "Scalable Bayesian inference for self-excitatory stochastic processes applied to big
American gunfire data." Statistics and Computing 31.1 (2021): 1-15.

2. Holbrook, Andrew J., Xiang Ji, and Marc A. Suchard. "Bayesian mitigation of spatial coarsening for a Hawkes model applied to gunfire, wildfire and viral contagion." Annals of Applied Statistics (2021), in press.

3. Holbrook, Andrew J., Xiang Ji, and Marc A. Suchard. "From viral evolution to spatial contagion: a biologically modulated Hawkes model." arXiv preprint arXiv:2103.03348 (2021).