Panagiotis (Panos) Toulis
Assistant Professor in Econometrics and Statistics
- ♦ nonparametric causal inference in complex systems through resampling/randomization methods;
e.g., randomization inference for causal peer effects, spillovers, interference in networks or integrated markets.
♦ statistical machine learning and inference using stochastic gradient
descent (SGD) with stable implicit updates.
Randomization tests of causal effects with interference between units, with Basse, G; Feller, A.
Biometrika, 2018, forthcoming
Stochastic gradient methods for estimation with large datasets, with Tran, D.; Airoldi, EM.
Journal of Statistical Software, 2018, forthcoming
Convergence diagnostics for stochastic gradient descent with constant step size, with Chee, J.
AI and Statistics, 2018 (AISTATS'18)
Asymptotic and finite-sample properties of estimators based on stochastic gradients, with Airoldi, EM.
Annals of Statistics, Volume 45, Number 4 (2017), 1694-1727