This is the second talk in RSSNI's Autumn Seminar Session.
Data to Dynamics with an application to modelling covid19
by Dr. Michelle Carey, UCD, Dublin, Ireland.
SEIR (Susceptible-Exposed-Infectious-Removed) models have been widely used to analyze the covid19 epidemic trend. These models facilitate a causal explanation for the drivers and impediments of the epidemic trend. But do they describe the behaviour of observed data? And how can we quantify the models' parameters that cannot be measured directly?
In this talk, I will introduce a methodology for estimating the solution; and the parameters of dynamical systems from incomplete and noisy observations of the processes. Building on the parameter cascading approach attributable to Ramsay et al. (2007), where a linear combination of basis functions approximates the implicitly defined solution of the dynamical system. The systems’ parameters are estimated so that this approximating solution adheres to the data. Comparing our approach with popular methods for estimating the parameters of dynamical systems, namely, the two-stage method, the nonlinear least-squares approach, parameter cascading and smooth functional tempering reveals an improved bias and sampling variance.
Michelle Carey, UCD, Dublin, Ireland