Interpretability of Epidemiological Models : The Curse of Non-Identifiability
This work addresses interpretability challenges for public health practitioners using epidemiological models, but it appears incremental as it refines existing concepts without introducing a new paradigm.
The paper tackles the problem of interpretability in epidemiological models by linking it to parameter identifiability, defining three notions to explore roles of model definition, loss function, fitting methodology, and data quality, and highlights these issues and mitigations within a compartmental model framework.
Interpretability of epidemiological models is a key consideration, especially when these models are used in a public health setting. Interpretability is strongly linked to the identifiability of the underlying model parameters, i.e., the ability to estimate parameter values with high confidence given observations. In this paper, we define three separate notions of identifiability that explore the different roles played by the model definition, the loss function, the fitting methodology, and the quality and quantity of data. We define an epidemiological compartmental model framework in which we highlight these non-identifiability issues and their mitigation.