Spatial-Temporal Modeling of COVID-19 Mortality Risk in Toronto, Canada


Thursday, April 29, 2021

Online Webinar & Q&A
Livestream Link –

Q&A – event code #35909

Dr. Cindy Feng Dr. Cindy Feng, Dalhousie University

Abstract: This talk will present a spatial-temporal model for modeling geo – referenced COVID -19 mortality data in Toronto, Canada. A range of factors and spatial-temporal terms are incorporated into the model. The non-linear and interactive effects of the neighborhood-level factors, i.e., population density and the average of income, are modeled as a two-dimensional spline smoother. Tensor product smoother is used for modeling the space-time interaction. By fitting this model, the residual spatial terms can provide insight into detecting high-risk areas not explained by the covariates. The predictive accuracy of the proposed model is evaluated based on out of sample predictive checking, and the findings showed that the model has high predictive power (10-fold cross-validation AUC =0.963) for predicting mortality risk among positive COVID -19 cases.

Bio: Dr. Cindy Feng received her MSc and PhD in Statistics from the Department of Statistics and Actuarial Science at Simon Fraser University in 2011. She is currently an Associate Professor at the Department of Community Health and Epidemiology, Faculty of Medicine at the Dalhousie University. Her research interests lie primarily in developing statistical models for analyzing correlated data in which repeated measurements, hierarchical clustering, multiple outcome types, and spatially correlated data might occur. Dr. Feng’s research has been funded by NSERC Discovery Grant, Canadian Statistical Sciences Institute, and MITACS. She has a deep desire to bridge the gap between statistical methods and practice by pursuing methodological development and applying statistical methods in public health, which has led her to develop partnerships with many researchers from various disciplines, i.e., medicine, psychology, biology, and sociology.