Meeting 2021
On November 15-16th 2021, Maastricht University’s School of Business and Economics and Copenhagen Business School jointly hosted the Causal Data Science Meeting 2021.
About
The two-day conference was once again held virtually and brought together researchers and experts from academia and industry, but this year with a growing number of participants and a larger line-up of speakers. More than 1,200 participants registered this year and the lineup included 34 speakers from leading universities such as Stanford, MIT, Harvard, Columbia University, Yale, and University of Amsterdam, as well as data scientists from industry such as Uber, Booking.com, Spotify, Toyota Research and Albert Heijn.
A summary of the event can be found here.
Keynotes
Sara Magliacane
Assistant Professor at the University of Amsterdam
She received her PhD at the VU Amsterdam on logics for causal inference under uncertainty in 2017, focusing on learning causal relations jointly from different experimental settings, especially in the case of latent confounders and small samples. After a year in IBM Research NY as a postdoc, she joined the MIT-IBM Watson AI Lab in 2019 as a Research Scientist, where she has been working on methods to design experiments that would allow one to learn causal relations in a sample-efficient and intervention-efficient way. Her current focus is on causality-inspired machine learning, i.e. applications of causal inference to machine learning and especially transfer learning, and formally safe reinforcement learning.
Guido Imbens
Professor at Stanford University
He has held tenured positions at UCLA, UC Berkeley, and Harvard University before joining Stanford in 2012. Imbens specializes in econometrics, and in particular methods for drawing causal inferences from experimental and observational data. He has published extensively in the leading economics and statistics journals. Together with Donald Rubin he has published a book, ”Causal Inference in Statistics, Social and Biomedical Sciences”. Guido Imbens is a fellow of the Econometric Society, the Royal Holland Society of Sciences and Humanities, the Royal Netherlands Academy of Sciences, the American Academy of Arts and Sciences, and the American Statistical Association. He holds an honorary doctorate from the University of St. Gallen. In 2017 he received the Horace Mann medal at Brown University, in 2021 he was awarded the Nobel Memorial Prize in Economic Sciences, jointly with David Card and Joshua Angrist. Currently Imbens is Editor of Econometrica.
Program
November 15, 2021
| Time | Presentation |
|---|---|
| 10:30 | Welcome Paul Hünermund, Jermain Kaminski, Carla Schmitt, Beyers Louw Copenhagen Business School & Maastricht University |
| Session 1 | |
| 10:40 | Self-fulfilling Bandits: Endogeneity spillover and dynamic selection in algorithmic decision-making Xiaowei Zhang Hongkong University |
| 10:55 | Off-policy learning of dynamic content promotions Joel Persson ETH Zürich |
| 11:10 | Estimating returns to special education: Combining machine learning and text analysis to address confounding Aurélien Sallin St. Gallen University |
| 11:25 | What’s on the telly? Causality for recommender systems in public-service media corporations Jordi Mur University of Barcelona |
| 11:40 | Q & A |
| 12:00 | 60 min break (Timer) |
| Session 2 | |
| 13:00 | Structural causal models are (solvable by) credal networks Alessandro Antonucci Dalle Molle Institute for Artificial Intelligence Research (IDSIA) |
| 13:15 | Estimating the probabilities of causation via deep monotonic twin networks Ciarán Lee Spotify Research |
| 13:30 | Double machine learning for sample selection models Martin Huber University of Fribourg |
| 13:45 | Positivity violation detection and explainability Hanan Shteingart Vian.ai |
| 14:00 | Q & A |
| 14:20 | 30 min break (Timer) |
| Session 3 | |
| 14:50 | Retrospective causal inference via matrix completion, with an evaluation of the effect of European integration on cross-border employment Jason Poulos Harvard Medical School |
| 15:05 | Crime and mismeasured punishment: Marginal treatment effect with misclassification Vitor Possebom Yale University |
| 15:20 | When should we (not) interpret linear IV estimands as LATE? Tymon Sloczynski Brandeis University |
| 15:35 | Preferences and productivity in organizational matching: Theory and empirics from internal labor markets Bo Cowgill Columbia Business School |
| 15:50 | Q & A |
| 16:10 | 30 min break (Timer) |
| Session 4 | |
| 16:40 | Experimentation and startup performance: Evidence from A/B testing Rem Koning Harvard Business School |
| 16:55 | The paper of how: Estimating treatment effects using the front-door criterion Marc Bellemare University of Minnesota |
| 17:10 | Causal-driven machine learning at Uber scale: A case study Okke van der Wal Uber |
| 17:25 | Generalizing experimental results by leveraging knowledge of mechanisms Carlos Cinelli University of Washington |
| 17:40 | Q & A |
| 18:00 | 30 min break (Timer) |
| Keynote | |
| 18:30 | Keynote Sara Magliacane University of Amsterdam & MIT-IBM Watson AI Lab |
November 16, 2021
| Time | Presentation |
|---|---|
| 10:30 | Welcome Paul Hünermund, Jermain Kaminski, Carla Schmitt, Beyers Louw Copenhagen Business School / Maastricht University |
| Session 1 | |
| 10:40 | The impact of the #MeToo movement on language at court: A text-based causal inference approach Henrika Langen University of Fribourg |
| 10:55 | Firm incentives and consumer adaption in a multi unit auction Simon Schulten Düsseldorf Institute for Competition Economics (DICE) |
| 11:10 | Drawing (causal) conclusions from data – some evidence Karsten Lübke FOM University of Applied Sciences |
| 11:25 | End-to-end causal analysis in Python with cause2e Daniel Gruenbaum Osram |
| 11:40 | Q & A |
| 12:00 | 60 min break (Timer) |
| Session 2 | |
| 13:00 | The role of the propensity score in fixed effect models Dmitry Arkhangelsky Center for Monetary and Financial Studies (CEMFI) |
| 13:15 | The challenges of measuring the impact of interventions in brick-and-mortar stores Patrick de Oude Albert Heijn |
| 13:30 | Longitudinal symptomatic interactions in long-standing schizophrenia: a novel five-point analysis based on directed acyclic graphs Giusi Moffa University of Basel |
| 13:45 | Causal inference with proxy variables in Booking.com Christina Katsimerou Booking.com |
| 14:00 | Q & A |
| 14:20 | 30 min break (Timer) |
| Session 3 | |
| 14:50 | Treatment effects in strategic management: with an application to choosing early stage venture capital Jorge Guzman Columbia Business School |
| 15:05 | Deep learning for individual heterogeneity: An automatic inference framework Max Farrell The University of Chicago Booth School of Business |
| 15:20 | Causal knowledge graph: A demonstration Victor Chen University of North Carolina at Charlotte |
| 15:35 | An empirical analysis of intra-firm product substitutability in fashion retailing Nathan Yang Cornell University |
| 15:50 | Q & A |
| 16:10 | 30 min break (Timer) |
| Session 4 | |
| 16:40 | The importance of being causal Iavor Bojinov Harvard Business School |
| 16:55 | Desiderata for representation learning: A causal perspective Yixin Wang University of Michigan |
| 17:10 | What experimental designs justify two-way-fixed-effects regression estimators Lihua Lei Stanford University |
| 17:25 | CausalML: A Python package for uplift modeling and causal inference with machine learning Zhenyu Zhao, Totte Harinen Tencent, Toyota Research |
| 17:40 | Q & A |
| 18:00 | 30 min break (Timer) |
| Keynote | |
| 18:30 | Keynote A Design Approach to Synthetic Controls Guido Imbens Stanford University |
Sponsors
The Causal Data Science Meeting 2021 was sponsored by Dataiku, Vianai and Vinted. Thank you for your support of the event!
Dataiku
Dataiku’s mission is big: to enable all people throughout companies around the world to use data by removing friction surrounding data access, cleaning, modeling, deployment, and more.
Vianai Systems
Vianai Systems, Inc. is a Human-Centered AI platform and products company launched in 2019 to address the unfulfilled promise of enterprise AI.
Vinted
Vinted is the largest online C2C marketplace in Europe dedicated to second-hand fashion with a growing community of 45+ million members spanning 15 markets.