Stat Cafe - Michael Price
Stationary Point Constrained Inference via DIffeomorphisms
- Time: Thursday, 2/5/2026 from 2:10PM to 3:30PM
- Location: BLOC 457
Description
Stationary points or derivative zero crossings of a regression function correspond to points where a trend reverses, making their estimation scientifically important. In our work, we develop a principled framework for functions with multiple regions of monotonicity by representing each function as the composition of a simple template and a smooth bijective transformation. This construction guarantees the estimated function will have a prespecified number of stationary points and enables joint inference on their locations. An application to brain signal estimation demonstrates the method’s accuracy and interpretability.
Our Speaker
Michael Price is a third-year PhD student in the Department of Statistics at Texas A&M University. He received a B.S. in Statistics from Rice University and worked as a Machine Learning Engineer in the financial tech industry for two years before coming to A&M. His research is generally focused on computational methods for learning complex Bayesian models, including nonparametric methods and models that arise from non-traditional datasets, such as time series and functional data.