Jason M. Klusowski


I am an assistant professor in the Department of Operations Research & Financial Engineering at Princeton University. Previously, I was an assistant professor in the Department of Statistics at Rutgers University, New Brunswick. I received my PhD in Statistics and Data Science from Yale University in 2018, where I was fortunate to have been mentored by Andrew Barron. From 2017 to 2018, I was a visiting graduate student in the statistics department at The Wharton School of the University of Pennsylvania.

My research is currently supported by NSF CAREER DMS-2239448 (PI) and formerly by NSF DMS-2054808 (PI) and TRIPODS DATA-INSPIRE Institute CCF-1934924 (Senior Personnel). Thank you, NSF!

I am a participating faculty at the Center for Statistics and Machine Learning (CSML) at Princeton and a senior personnel on TRIPODS DATA-INSPIRE Institute at Rutgers.

My spouse is an assistant professor of marketing at Yale University. 


  • February 2024. I'm looking for a postdoctoral associate in statistics and machine learning at Princeton, starting summer/fall 2024!
  • February 2024. I received an Innovation Research Grant from the School of Engineering and Applied Science (SEAS) at Princeton.
  • January 2024. Paper on convergence rates of oblique regression trees with Matias Cattaneo and Rajita Chandak to appear in Annals of Statistics.
  • December 2023. New paper on stochastic gradient descent for additive nonparametric regression with Xin Chen.
  • October 2023. New paper on inference with Mondrian random forests with Matias Cattaneo and William Underwood.
  • October 2023. New paper on robust transfer learning with Jianqing Fan and Cheng Gao.
  • September 2023. New paper on stacked regressions with Xin Chen and Yan Shuo Tan.
  • September 2023. New paper on the implicit bias of Adam with Matias Cattaneo and Boris Shigida.

Research Interests

I am broadly interested in statistical machine learning for complex, large scale models. My work seeks to describe the tensions among interpretability, statistical accuracy, and computational feasibility.

  • decision trees and ensemble learning (CART, random forests, stacking)
  • neural networks (approximation theory and statistical properties)
  • gradient-based optimization (ADAM, SGD)
  • large limit behavior of statistical models (Lasso, Slope)