Spatio-temporal data modeling with applications to weather and disease

Meteorological and epidemiological data are oftentimes collected over many years at various locations. In such cases, it is beneficial to use spatio-temporal modeling to account for trends and the correlation of nearby observations. This thesis explores applications to spatio-temporal modeling. First, a method is developed to model the marginal distribution of spatial extreme values at a large scale quickly while allowing flexibility by introducing a fused penalty for parameter regularization. Next, various models are considered and evaluated to compare county-level HIV prediction over the US to determine if spatial models are advantageous when an abundance of covariates are available that capture the data variability. Lastly, a generalized additive model with spatial and temporal covariates is utilized to evaluate the impact of adulticide spraying on gravid Culex mosquitoes in the North Shore Mosquito Abatement District of Illinois.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Statistics

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2021

Author dc:creator

• Sass, Danielle

Contributors dc:contributor

• Li, Bo
• Simpson, Douglas
• Douglas, Jeffrey
• Park, Trevor

• adulticide spray
• fused penalties
• HIV prediction
• spatial extremes
• West Nile virus

Statement dc:rights

• Copyright 2021 Danielle Sass

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/110813

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/110813