Jose Blanchet (Stanford)

Event Date: 

Wednesday, March 14, 2018 - 3:30pm

Event Date Details: 

Refreshments at 3.15 pm

Event Location: 

  • Sobel Room (SH 5607F)
Title: Distributionally Robust Optimization via Optimal Transport and Its Applications
Optimal mass transportation is a powerful tool in the arsenal of many quantitative disciplines, with well documented applications spanning a wide range of areas, including, operations research, economics and image analysis. In this talk, we focus on data-driven distributionally robust optimization, that is, a class of perfect-information games in which an optimizer selects an action and adversary chooses a model within a region around a baseline distribution, which we often take to be an empirical measure. We show how many machine learning algorithms can be retrieved as a special cases of this type of formulation. We establish connections to regularized portfolio optimization strategies that are common in practice. These connections provide a rich intuition which allows to interpret various regularization parameters which are typically chosen in practice via cross validation, but owing to this intuition, we are able to define a reasonable optimization criterion for choosing regularization parameters via pivotal statistics, thereby avoiding time-consuming cross-validation.

(This talk is based on joint work with Yang Kang, Karthyek Murthy and Fan Zhang).

Here are two papers which are the basis for the talk: