Statistics is the backbone of experimental and observational science, and as such it appears in most quantitative fields; it also give us many tools to understand machine learning. With colleagues, I try to distill important concepts, and develop pedagogically sound papers, software and courses. Currently I’m writing chapters for the upcoming textbook Bayesian Workflow; an early outline of this work exists as a preprint.

I have now several times taught 1 – 2 days workshops on modeling and data analysis. My courses cover fundamental concepts in Bayesian statistics and some of the motivating theory, demonstrations on scientific examples, discussions of the “under-the-hood” algorithms that support statistical software, and hands-on coding exercises. Here’s a testimony on a workshop I gave in 2020 at McGill University in Montreal. I quite enjoy teaching workshops on Stan and Torsten. If you’re interested in such a course, feel free to contact me!

As a graduate student, I also served as a teacher assistant for five consecutive years for the department of statistics at Columbia university, for courses at all levels (undergraduate, masters and PhD). Upon graduation, I recieved the Minghui Yu Teaching Assistant Award, awarded by the director of graduate studies based on feedback from students.


I’ve been invited as an intrusctor for the following courses and conferences:

  • (2018 – 2022) Introduction to Probability and Bayes, PHC 506: Biometry in Pharmaceutics, University of Buffalo, School of Pharmacy, Buffalo, NY. [notes, erratum]
  • (2019 – 2022) Building, fitting, and criticizing Bayesian Pharmacokinetic/Pharmacodynamic models using Stan and Torsten , University of Buffalo, School of Pharmacy, Buffalo, NY. (1 day workshop) [outline]
  • (2019, 2020) Stan for the people: introductory workshop on Bayesian modeling, McGill University, Montreal, Canada. (2 days workshop)
  • (2019) Population and ODE-based models using Stan and Torsten, co-instructor with Yi Zhang, Stan Conference 2019, Cambridge, UK
  • (2018) How to Develop for the Stan C++ Core Language, Stan Conference 2018, Pacific Grove, CA. [slides]
  • (2017) Introduction to Bayesian Data Analysis with Stan, STAT 220: Bayesian Statistics, Harvard University, Cambridge, MA

Teacher assistant

Courses at Columbia:

  • (2021, 2022) STAT 6102: Applied Statistics II (PhD level)
  • (2021) STAT 4206 / 5206: Statistical Computing and Introduction to Data Science (undergraduate and masters level)
  • (2019, 2020) STAT 6701: Foundations of Grapical Models (PhD level))
  • (2019) STAT 4204/5204: Statistical inference (undergraduate / masters level)
  • (2018) STAT 4206: Statistical Computing and Introduction to Data Science (undergraduate level)
  • (2018) STAT 5224: Bayesian Statistics (masters level)
  • (2017) STAT 4205: Linear Regression Models (undergraduate level)


  • (2017) Stan for Physics, Massachusetts Institute of Technology, Cambridge, MA (5 days workshop, taught by Michael Betancourt)
  • (2017) Getting Started with Bayesian PKPD Modeling using Stan and Torsten, Population Approach Group in Europe 26, Budapest, Hungary (1 day workshop, taught by Bill Gillespie)
  • (2016, 2017) Getting Started with Bayesian PKPD Modeling using Stan, American Conference on Pharmacometrics 6 and 7 (1 day workshop, taught by Bill Gillespie)


Occasionally I write posts on the blog Statistical modeling, causal inference, and social science, primarily curated by Andrew Gelman, in an effort to summarize ongoing research: