Short Video Contest Winner: Optimization of a Metallurgical Process with Uncertain Dynamics
It’s been a busy summer and early fall. Earlier in the year I submitted an entry to the Society of Mining, Metallurgy & Exploration Mineral Processing Division Short Video Contest titled Optimization of a Metallurgical Process with Uncertain Dynamics. Entries consisted of a one page abstract and a less than three minute video presenting the work.
Optimization of a Metallurgical Process with Uncertain Dynamics provided an example of how Gaussian process regression can be used with Bayesian optimization to find the optimum of a chemical process for rare earth element extraction in few experiments.
This application of probabilistic optimization won first place price in the contest based on both the content of the abstract and video submissions. Instead of going into more details here, the video and abstract probably do a better job.
Scott Koermer
Post-Doctoral Researcher
Post doctoral researcher at Los Alamos National Lab with the statistical sciences group. During my time at Los Alamos I have worked on explosion monitoring applications as well as model/prior selection for Bayesian Neural Networks. Much of my research in the past 6 years has focused on applying Bayesian methods to scientific and engineering applications, including modeling processes for optimizing the recovery of critical minerals from waste streams.