Department of Applied Mathematics, University of Waterloo
Identifiability and Sensitivity Analysis
When working with models, it is important that we develop models that are robust and accurate in describing the phenomena in question. To do this, we can analyze the model's identifiability and its sensitivity to the model parameters. Identifiability describes the ability to acquire parameters to accurately describe a given set of experimental data. Sensitivity analysis, on the other hand, describes the effect of the input parameters or factors on a given model output. This seminar will discuss both structural and practical identifiability, as well as parameter estimation and (local and global) sensitivity analysis. More specifically, methods will be presented to assess and analyze models with respect to these factors.