معرفی
Panagiotis Toulis (Panos) serves as Associate Professor of Econometrics and Statistics and John E. Jeuck Faculty Fellow at the University of Chicago Booth School of Business, where his research focuses on model-agnostic causal inference methods for complex networked environments using randomization techniques that provide robustness beyond classical statistical approaches.
His academic credentials include:
- Ph.D. in Statistics from Harvard University (2016) under Edo Airoldi, David Parkes, and Don Rubin
- Master’s in Computer Science from Harvard University (2011) under David Parkes
- M.Eng. in Electrical and Computer Engineering from Aristotle University of Thessaloniki, Greece (2006)
Toulis pioneers randomization and permutation tests for causal inference under network interference, developing methods applicable to experimental design in complex domains where traditional model-based approaches fail. His work bridges statistics, optimization, and economics, with significant contributions to invariance-based inference for observational studies and machine learning integration through stochastic gradient descent frameworks. This research has practical applications in business analytics, vaccine distribution, and economic policy.
Analysis of his 15 most recent publications (2023-2025) reveals dominant trends in extending randomization tests to handle network spillover effects, two-sided market experiments, and temporal dynamics with treatment habituation. A growing emphasis on machine learning assistance for complex treatment effect detection demonstrates his innovative fusion of statistical rigor with computational scalability, consistently providing finite-sample theoretical guarantees uncommon in asymptotic-focused literature.
Toulis has earned recognition through prestigious awards:
- Arthur P. Dempster Award from Harvard University’s Department of Statistics
- LinkedIn Economic Graph Challenge award
- 2012 Google United States/Canada PhD Fellowship in statistics
He teaches core Booth courses including Applied Regression Analysis (41100), Causal Inference for Business Applications (41207), and the Econometrics and Statistics Colloquium (41600), while maintaining active industry connections through prior software engineering roles at Google and Greek startups. His research appears in top journals including Journal of the Royal Statistical Society and Annals of Statistics, addressing critical methodological challenges in modern data science.
No dedicated laboratory structure is documented, but his work inherently involves collaborative research networks across statistics, economics, and computer science disciplines through the Booth School’s academic infrastructure.
