About
I'm a scientist at Verisk, where I build and validate statistical models of tropical cyclones and lend a hand with the risk from other atmospheric perils, like winter storms and severe thunderstorms.
Before that I was a research associate at Indiana University, developing Bayesian methods and open-source R software for analyzing fMRI data on the cortical surface. I earned my PhD in Statistical Science at UC Santa Cruz, where my dissertation on Bayesian tensor regression was a finalist for ISBA's Savage Award. Long before any of that, I taught high-school math in Portuguese as a Peace Corps volunteer in Zóbuè, Mozambique.
Interests
- Bayesian modelingHierarchical models, MCMC, and honest uncertainty quantification.
- Climate & catastrophe riskSimulating tropical cyclones and other perils over thousands of years.
- NeuroimagingSpatial models for task fMRI and tensor methods for brain data.
- Statistical softwareR packages, Rcpp and C++, and fast, reproducible analysis.
- Data visualizationClear, accurate plots. I taught a graduate course on it.
- TeachingFrom Mozambique to Penn State and UC Santa Cruz classrooms.
Software
- BayesfMRI · surface-based spatial Bayesian models for task fMRI (contributor)
- bayestensorreg · Bayesian tensor regression using tensor decompositions
- brainSim · simulating CIFTI brain data
Writing
A few posts on statistics and R, most recently Bayesian COVID Probability and Estimating Negative Variance. All posts →