Meet our Teams
Guelph Team Core Researchers
Dr. Monica Cojocaru
AI4Casting Hub Director, Associate Dean (Research & Grad Studies, CEPS), Professor of Mathematics and Statistics
Dr. Edward Thommes
Adjunct Professor and Lead, New Products and Innovation, Sanofi
Christopher Van Bommel
Postdoctoral Scholar
Rhiannon Loster
Public Health Ontario (PHO) Data Analyst
Darren Flynn-Primrose
Data Scientist
Benjamin Benteke Longaou
PhD Candidate
Siddhesh Suresh Kadam
AI4Casting Hub Administrator
Daniel Dombrovsky
Website Developer
Affiliated Teams
University of British Columbia
Daniel J. McDonald
Associate Professor, Department of Statistics, UBC
Christine Chuong
MSc Student, UBC
Forecasting Submission Teams
University of Guelph Dynamics Training Lab
Contributors
Christopher van Bommel, Edward Thommes & Monica Cojocaru
Model Description
The composite curve model forecasts vaccine uptake by correlating historical Google search trends with past vaccination rates. It identifies the lag between peak search activity and vaccination uptake, using this relationship to predict future vaccination behaviors.
Johns Hopkins University Applied Physics Lab
Contributors
William T. Redman & Luke Mullany
Model Description
A sliding window is used to construct an temporally local approximation of the Koopman operator. Time-delays and radial basis functions are used for the lifting to function space. All states (and D.C. + P.R.) are used to construct the approximation of the Koopman operator to estimate couplings/correlations. Gaussian noise, of variance estimated from previous years, is used to sample possible variants of the data to approximate the forecast percentiles.
University of British Columbia Statistics
Contributors
Daniel McDonald & Christine Chuong
Model Description
Use target values from historic seasons in a window around the target week to form quantiles, subtracting off the median and adding location specific medians
Virginia Tech Sanghani Center for Artificial Intelligence and Data Analytics
Contributors
Yiqi Su, Patrick Butler & Naren Ramakrishnan
Model Description
This ensemble forecasts national-, provincial-, and regional-level laboratory detection of respiratory viruses, including SARS-CoV-2, RSV, and influenza. Included models focus on novel deep learning techniques.
