Optimization-Conscious Econometrics Conference
November
15-16
2019
Photos
Astonishing recent progress in numerical optimization presents many opportunities and challenges for modern data analysis and scientific inference. Econometrics and statistics offer a valuable angle from which to study optimization problems and, conversely, operations research offers a productive perspective from which to consider many problems in econometrics and statistics, whence the importance of fostering interactions between academic communities.
This conference brought together scholars from economics, applied mathematics, operations research, computer science and statistics to share their experience and expertise on these important issues.
ORGANIZERS
Agenda
Friday, November 15, 2019
8:00 am - 8:25 am
Registration and Breakfast
8:25 am - 8:30 am
Introductory Remarks
Session 1: Modern Optimization in Econometrics
8:30 am - 9:15 am
Nonparametric Maximum Likelihood Methods for Binary Response Models with Random Coefficients
Jiaying Gu, University of Toronto
9:15 am - 10:00 am
L0 Penalized Quantile Regression
Simon Lee, Columbia University
10:00 am - 10:45 am
Nonlinear Earnings and Employment Dynamics at the Extensive and Intensive Margins
Manuel Arellano, CEMFI
10:45 am - 11:00 am
Break
11:00 am - 11:45 am
Deep Inference: Artificial Intelligence for Structural Model Estimation
Elena Manresa, New York University
11:45 am - 12:30 pm
Recovering Latent Variables by Matching
Stephane Bonhomme, University of Chicago
12:30 pm - 1:15 pm
Lunch
Session 2: Polynomial Optimization
1:15 pm - 2:00 pm
Polynomial Optimization and Symmetry
Annie Raymond, University of Massachusetts
2:00 pm - 2:45 pm
The Moment-SOS Hierarchy: with Applications in Optimization, Probability, Statistics, Control, Non-Linear PDEs, Computational Geometry
Jean-Bernard Lasserre, LAAS-CNRS, Toulouse
2:45 pm - 3:00 pm
Identification and Estimation of Dynamic Random Coefficient Models
Wooyong Lee, University of Chicago
3:00 pm - 3:30 pm
Break
Session 3: Questions at the Intersection of Optimization and Econometrics
3:30 pm - 4:15 pm
Algorithmic Sampling from an Econometric Perspective
Simon Lee, Columbia University
4:15 pm - 5:00 pm
Optimization, Diffusion, and Dimension Dependence
Michael Jordan, University of California, Berkeley
5:00 pm - 5:45 pm
Optimization in Statistics and Data Science: Some Perspectives Past and Present
Stephen Wright, University of Wisconsin
5:45 pm
Adjourn for the Day
Saturday, November 16, 2019
8:30 am - 9:00 am
Breakfast
Session 5: Integer Programming, Econometrics and Statistics
9:00 am - 9:45 am
Machine Learning Under a Modern Optimization Lens
Dimitris Bertsimas, MIT
9:45 am - 10:30 am
Structured Statistical Learning at Scale: Convex and Mixed Integer Optimization Perspectives
Rahul Mazumder, MIT
10:30 am - 10:45 am
Break
10:45 am - 11:30 am
Advances at the Intersection of Integer Programming, Data Science, and Econometrics
Andrea Lodi, Polytechnique Montréal
11:30 am - 12:15 pm
Building Representative Matched Samples with Multi-valued Treatments in Large Observational Studies
Jose Zubizarreta, Harvard University
12:15 pm - 1:00 pm
Lunch
Session 6: Linear Programming and Econometrics
1:00 pm - 1:45 pm
Vector Quantile Regression: Quantile Regression Meets Optimal Transport
Alfred Galichon, New York University
1:45 pm - 2:30 pm
Differentiable Ranks and Quantiles Using Optimal Transport
Marco Cuturi, Google / CREST - IPP
2:30 pm - 3:15 pm
Nonparametric Estimates of Demand in the California Health Insurance Exchange
Alexander Torgovitsky, University of Chicago
3:15 pm - 3:30 pm
Linear Programming Aspects of Regression Rankscore Inference
Yuehao Bai, University of Chicago
3:30 pm - 3:45 pm
Break
Session 7: Stochastic Programming and Econometrics
3:45 pm - 4:30 pm
Inference by Stochastic Optimization: A Free Lunch Bootstrap
Jean-Jacques Forneron, Boston University
4:30 pm - 5:15 pm
Spatial Price Integration in Competitive Markets with Capacitated Transportation Networks
John Birge, University of Chicago
5:15 pm - 5:20 pm