Prof. Panagiotis (Panos) Toulis
Panagiotis (Panos) Toulis

Panagiotis (Panos) Toulis
Assistant Professor of Econometrics and Statistics
Address: 5807 S Woodlawn Ave
Office: 358
tel: 773.834.5953


Working papers

  • PT, Alexander Volfovsky, Edoardo M. Airoldi, "Causal inference with entangled treatments" (2015)
  • PT, Edoardo M. Airoldi, Donald B. Rubin, "Causal inference under partially revealed interference" (2015)
  • PT, Edoardo M. Airoldi, "Implicit stochastic approximation" (2015,  pdf)
  • Dustin Tran, PT, Edoardo M. Airoldi, "Stochastic gradient methods for estimation with large datasets" (2015,   pdf, minor revision)

Journal papers

  • PT, Edoardo M. Airoldi, "Asymptotic and finite-sample properties of estimators based on stochastic gradients", Annals of Statistics, 2017, forthcoming (  pdf)
  • PT, "A useful pivotal quantity", American Statistician, 2016, forthcoming (   www |   pdf |    bib)
  • PT, Edoardo M. Airoldi, "Scalable estimation strategies based on stochastic approximations: Classical results and new insights", Statistics and Computing, 2015 (  www |   pdf |   bib).
  • PT, David Parkes, "Design and analysis of multi-hospital kidney-exchanges using random graphs", Games and Economic Behavior, 2015 (  pdf |   www |   bib)

Conference papers

  • PT, David Parkes, "Long-term causal effects via behavioral game theory", Neural Information Processing Systems, 2016, Barcelona, Spain, (NIPS'16,   www |   bib)
  • PT, Dustin Tran, Edoardo M. Airoldi, "Towards stability and optimality in stochastic gradient descent" , AI and Statistics, 2016, Cadiz, Spain (AISTATS' 16,   pdf |   bib)
  • PT, David Parkes, Elery Pfeffer, James Zou, "Incentive-compatible experimental design", Economics and Computation, 2015, Portland, Oregon (EC'15,    pdf |   bib)
  • PT, Jason Rennie, Edoardo Airoldi, "Statistical analysis of stochastic gradient methods for generalized linear models", International Conference of Machine Learning, 2014, Beijing, China (ICML' 14,   www,   pdf |   bib)
  • PT, Edward Kao, "Estimation of Causal Peer Influence Effects", International Conference of Machine Learning, 2013, Atlanta, Georgia (ICML'13,   www,   pdf |   bib)
  • PT, David Parkes, "A Random Graph Model of Kidney Exchanges: Efficiency, Individual-Rationality and Incentives", Economics and Computation, 2011, San Jose, California (EC'11,   pdf |   bib)
  • Nikolaos Mavridis, Wajahat Kazmi, PT, C. Ben-AbdelKader, "On the synergies between online social networking, Face Recognition, and Interactive Robotics", International Conference on Computational Aspects of Social Networks, 2009, Fontainebleau, France (CASoN'09,   pdf |   bib)
  • PT, Dionisis Kehagias, Pericles Mitkas, "Mertacor, a successful trading agent", Autonomous Agents and Multi-Agent Systems, 2006, Hakodate, Japan (AAMAS'06,   pdf |   bib)
  • Dionisis Kehagias, PT, Pericles Mitkas, "A Long-Term Profit Seeking Strategy for Continuous Double Auctions in a Trading Agent Competition", Fourth Hellenic Conference on Artificial Intelligence, 2006, Heraklion, Crete (  www |   pdf |   bib)

Book chapters

  • PT, Edoardo Airoldi, "Stochastic gradient methods for principled estimation with large datasets", in "Handbook of Big Data", CRC Press, 2016 (eds. Buhlmann et. al.)-(   www |   pdf )
  • Nikolaos Mavridis, Wajahat Kazmi, PT, "Friends with Faces: How Social Networks Can Enhance Face Recognition and Vice Versa", Computational Social Network Analysis (eds. A. Ajith & H. Aboul-Ella & S. Vaclav), pp. 453-482, Springer London

Short papers, Workshops, and Tutorials

  • Introduction to Stochastic Gradient Descent. This is a short introduction to Stochastic Gradient Descent, trying to cover both the optimization and statistical perspective. It covers classical literature in stochastic approximation, as well as recent developments.
  • Matt Bonakdarpour, PT, "Statistical perspectives of stochastic optimization" (NIPS'16, Probabilistic Numerics Workshop,   pdf)
  • Implicit temporal differences (  pdf) - Using ideas from implicit stochastic gradient descent to improve the stability of the Temporal Differences algorithm (TD) in reinforcement learning. Neural Information and Processing Systems (NIPS 2014, Montreal, Canada).
  • Software Engineering With R, 2013 (  pdf) - Intro to software engineering practices with R: unit testing, debugging, logging, profiling.
Booth Homepage | Booth Intranet | UC Homepage Copyright © 2016 Chicago Booth