Parameter estimation on a toy physics problem
The goal of this talk is to introduce a Bayesian approach to parameter estimation and apply this approach to a toy physics problem. In this problem, we have data that transitions from one steady state to another. We will discuss how using more data in the parameter estimation process might not always be the best idea; specifically, using more steady state data lessens the efficiency of the parameter estimation. We will lastly discuss results from the toy problem which will allow us to determine the appropriate amount of steady state versus transition state data and how this might be applied to more challenging inverse problems.
(Bagels and refreshments will be served.)