Event Category: Operations Research Seminar Series
-
Sparsity, Feature Selection and the Shapley Folkman Theorem
10/8/20 | 4:15pm | Online only Alexandre d’Aspremont Professor ENS Abstract: The Shapley Folkman theorem acts a central limit theorem for convexity: It shows that Minkowski sums of arbitrary bounded sets are increasingly close to their convex hull as the number of terms in the sum increases. This produces a priori…
-
Contextual Optimization: Bridging Machine Learning and Operations
10/1/20 | 4:15pm | Online only Adam Elmachtoub Assistant Professor Columbia Abstract: Many operations problems are associated with some form of a prediction problem. For instance, one cannot solve a supply chain problem without predicting demand. One cannot solve a shortest path problem without predicting travel times. One cannot solve a…
-
Sparse Regression at Scale
9/10/20 | 4:15pm | Online only Hussein Hazimeh ORC PhD Student MIT Abstract: We consider the least squares regression problem, penalized with a combination of the L0 and L2 norms (a.k.a. L0L2 regularization). Recent work presents strong evidence that the resulting L0-based estimators can outperform popular sparse learning methods, under many…
-
Information and Memory in Dynamic Resource Allocation
2/27/20 | 4:15pm | E51-325 Reception to follow. Kuang Xu Associate Professor Stanford Abstract: We propose a general framework, dubbed Stochastic Processing under Imperfect Information (SPII), to study the impact of information constraints and memories on dynamic resource allocation. The framework involves a Stochastic Processing Network (SPN) scheduling problem in which the…
-
Online Allocation and Pricing: Constant Regret via Bellman Inequalities
9/17/20 | 4:15pm | Online only Itai Gurvich Professor Cornell Abstract: In this talk, I will discuss a family of dynamic resource allocation problems that includes, as specific instances, network revenue management and dynamic pricing. I will show that the value of information (or the regret) — the expected gap in…
-
On the Global Convergence and Approximation Benefits of Policy Gradient Methods
12/3/20 | 4:15pm | Online only Dan Russo Assistant Professor Columbia Abstract: Policy gradient methods apply to complex, poorly understood, control problems by performing stochastic gradient descent over a parameterized class of polices. Unfortunately, due to the multi-period nature of the objective, they face non-convex optimization problems. The first part of the…
-
Minimum Earnings Regulation and the Stability of Marketplaces
11/19/20 | 4:15pm | Online only Garrett Van Ryzin Distinguished Scientist Amazon Abstract: We build a model to study the implications of utilization-based minimum earning regulations of the kind enacted by New York City (and recently Seattle) for its ride-hailing providers. We identify the precise conditions under which a utilization-based minimum…
-
Personalizing Treatments Using Microbiome and Clinical Data
10/29/20 | 4:15pm | Online only Eran Segal Professor Weizmann Institute of Science Abstract: Accumulating evidence supports a causal role for the human gut microbiome in obesity, diabetes, metabolic disorders, cardiovascular disease, and numerous other conditions. I will present our research on the role of the human microbiome in health and…
-
Robust Convex Optimization: A New Perspective That Unifies And Extends
9/24/20 | 4:15pm | Online only Dick den Hertog Professor University of Amsterdam Abstract: Robust convex constraints are difficult to handle, since finding the worst-case scenario is equivalent to maximizing a convex function. In this paper, we propose a new approach to deal with such constraints that unifies all approaches known…
-
Superhuman AI for Multiplayer Poker
3/5/20 | 4:15pm | E51-325 Reception to follow. Tuomas Sandholm Professor Carnegie Mellon Abstract: In recent years there have been great strides in AI, with games often serving as challenge problems, benchmarks, and milestones for progress. Since the 1950s, poker has served as such a challenge problem in AI, game theory,…