SSA QLD Branch Meeting: From uncertainty to certainty: Exploiting parameter relationships to manage uncertainty in mathematical modelling

Please join us online or in person for our October Queensland Branch Meeting. The seminar will start at 5:00 pm. Details for the seminar are provided below.

TITLE: From uncertainty to certainty: exploiting parameter relationships to manage uncertainty in mathematical modelling

SPEAKER: Gloria Milena Monsalve Bravo, University of Queensland

TIME: 5:00 - 6:30 pm (AEST), 18^th October 2023

VENUE:  0500D-512 - 88 Creek Street, Brisbane and online (Zoom details will be sent with registration).
Special instructions for in-person venue: Enter through the main door at 308 Queen Street and pass through the Atrium. Speak to the concierge at the elevator located at the back of the room, and notify them that you are attending the Statistical Society of Australia event.

Please note that the seminar will be recorded and might be put on YouTube or similar platform.

ABSTRACT:

Deterministic mathematical models are widely used across ecology, biology, and chemistry to interrogate the mechanisms that underpin natural and physical processes. However, as our understanding of these processes improves, models are made more complex and often require many more parameters to be estimated from available data. Unfortunately, in many of these mathematical models, so-called “sloppy”, parameters cannot be uniquely estimated, and so a significant amount of uncertainty in parameter values often remains after even a very successful fit of the model to data. The reason for this is that the model’s behavior is often controlled by a relatively small number of stiff parameter combinations which significantly influence model predictions. This talk, therefore, focuses on how to identify these key parameter combinations by using a Bayesian inference-based approach to assess the sensitivity of model outputs to changes in parameter values, constrained by the combination of both data and prior beliefs (e.g., from experiments and/or expert knowledge). In this way, this approach can reveal which of the key parameter combinations are informed primarily by the data or are also substantively influenced by the priors. Different examples in ecology, biology, and chemistry will be discussed to showcase the benefits of this technique for applications where mathematical models need to be fitted to data while focusing on the very common context in complex systems where the amount and quality of data are low compared to the number of model parameters to be collectively estimated. Furthermore, it will be shown how the stiff parameter combinations, once identified, can uncover controlling mechanisms underlying the system being modeled and also inform which of the model parameters need to be prioritized in future experiments for improved parameter inference from collective model-data fittings.

SPEAKER'S BIO: