Shreyas Sekar’s research while affiliated with Harvard University and other places

We study the canonical problem of maximizing a stochastic submodular function subject to a cardinality constraint, where the goal is to select a subset from a ground set of items with uncertain individual performances to maximize their expected group value. Although near-optimal algorithms have been proposed for this problem, practical concerns regarding scalability, compatibility with distributed implementation, and expensive oracle queries persist in large-scale applications. Motivated by online platforms that rely on individual item scores for content recommendation and team selection, we study a special class of algorithms that select items based solely on individual performance measures known as test scores. The central contribution of this work is a novel and systematic framework for designing test score–based algorithms for a broad class of naturally occurring utility functions. We introduce a new scoring mechanism that we refer to as replication test scores and prove that as long as the objective function satisfies a diminishing-returns condition, one can leverage these scores to compute solutions that are within a constant factor of the optimum. We then extend these scoring mechanisms to the more general stochastic submodular welfare-maximization problem, where the goal is to partition items into groups to maximize the sum of the expected group values. For this more difficult problem, we show that replication test scores can be used to develop an algorithm that approximates the optimal solution up to a logarithmic factor. The techniques presented in this work bridge the gap between the rigorous theoretical work on submodular optimization and simple, scalable heuristics that are useful in certain domains. In particular, our results establish that in many applications involving the selection and assignment of items, one can design algorithms that are intuitive and practically relevant with only a small loss in performance compared with the state-of-the-art approaches. This paper was accepted by Chung Piaw Teo, optimization.

We study the equilibrium behavior in a multi-commodity selfish routing game with uncertain users where each user over- or under-estimates their congestion costs by a multiplicative factor. Surprisingly, we find that uncertainties in different directions have qualitatively distinct impacts on equilibria. Namely, contrary to the usual notion that uncertainty increases inefficiencies, network congestion decreases when users over-estimate their costs. On the other hand, under-estimation of costs leads to increased congestion.

The increasingly tight coupling between humans and system operations in domains ranging from intelligent infrastructure to e-commerce has led to a challenging new class of problems founded on a well-established area of research: incentive design. There is a clear need for a new tool kit for designing mechanisms that help coordinate self-interested parties while avoiding unexpected outcomes in the face of information asymmetries, exogenous uncertainties from dynamic environments, and resource constraints. This article provides a perspective on the current state of the art in incentive design from three core communities—economics, control theory, and machine learning—and highlights interesting avenues for future research at the interface of these domains. Expected final online publication date for the Annual Review of Control, Robotics, and Autonomous Systems Volume 2 is May 3, 2019. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

In this work we study a problem that naturally arises in the context of several important applications, such as online dating, kidney exchanges, and team formation. Given an uncertain, weighted (hyper)graph, how can we efficiently find a (hyper)matching with high expected reward, and low risk? We introduce a novel formulation for finding matchings with maximum expected reward and bounded risk under a general model of uncertain weighted (hyper)graphs. Given that our optimization problem is NP-hard, we turn our attention to designing efficient approximation algorithms. For the case of uncertain weighted graphs, we provide a $\frac{1}{3}$-approximation algorithm, and a $\frac{1}{5}$-approximation algorithm with near optimal run time. For the case of uncertain weighted hypergraphs, we provide a $\varOmega (\frac{1}{k})$-approximation algorithm, where k is the rank of the hypergraph (i.e., any hyperedge includes at most k nodes), that runs in almost (modulo log factors) linear time.We complement our theoretical results by testing our algorithms on a wide variety of synthetic experiments, where we observe in a controlled setting interesting findings on the trade-off between reward, and risk. We also provide an application of our formulation for providing recommendations of teams that are likely to collaborate, and have high impact. Code related to this paper is available at: https://github.com/tsourolampis/risk-averse-graph-matchings.

The design of personalized incentives or recommendations to improve user engagement is gaining prominence as digital platform providers continually emerge. We propose a multi-armed bandit framework for matching incentives to users, whose preferences are unknown a priori and evolving dynamically in time, in a resource constrained environment. We design an algorithm that combines ideas from three distinct domains: (i) a greedy matching paradigm, (ii) the upper confidence bound algorithm (UCB) for bandits, and (iii) mixing times from the theory of Markov chains. For this algorithm, we provide theoretical bounds on the regret and demonstrate its performance via both synthetic and realistic (matching supply and demand in a bike-sharing platform) examples.

Renewable resources are starting to constitute a growing portion of the total generation mix of the power system. A key difference between renewables and traditional generators is that many renewable resources are managed by individuals, especially in the distribution system. In this paper, we study the capacity investment and pricing problem, where multiple renewable producers compete in a decentralized market. It is known that most deterministic capacity games tend to result in very inefficient equilibria, even when there are a large number of similar players. In contrast, we show that due to the inherent randomness of renewable resources, the equilibria in our capacity game becomes efficient as the number of players grows and coincides with the centralized decision from the social planner's problem. This result provides a new perspective on how to look at the positive influence of randomness in a game framework as well as its contribution to resource planning, scheduling, and bidding. We validate our results by simulation studies using real world data.

Renewable resources are starting to constitute a growing portion of the total generation mix of the power system. A key difference between renewables and traditional generators is that many renewable resources are managed by individuals, especially in the distribution system. In this paper, we study the capacity investment and pricing problem, where multiple renewable producers compete in a decentralized market. It is known that most deterministic capacity games tend to result in very inefficient equilibria, even when there are a large number of similar players. In contrast, we show that due to the inherent randomness of renewable resources, the equilibria in our capacity game becomes efficient as the number of players grows and coincides with the centralized decision from the social planner's problem. This result provides a new perspective on how to look at the positive influence of randomness in a game framework as well as its contribution to resource planning, scheduling, and bidding. We validate our results by simulation studies using real world data.

Renewable resources are starting to constitute a growing portion of the total generation mix of the power system. A key difference between renewables and traditional generators is that many renewable resources are managed by individuals, especially in the distribution system. In this paper, we study the capacity investment and pricing problem, where multiple renewable producers compete in a decentralized market. It is known that most deterministic capacity games tend to result in very inefficient equilibria, even when there are a large number of similar players. In contrast, we show that due to the inherent randomness of renewable resources, the equilibria in our capacity game becomes efficient as the number of players grows and coincides with the centralized decision from the social planner's problem. This result provides a new perspective on how to look at the positive influence of randomness in a game framework as well as its contribution to resource planning, scheduling, and bidding. We validate our results by simulation studies using real world data.

Design of incentives or recommendations to users is becoming more common as platform providers continually emerge. We propose a multi-armed bandit approach to the problem in which users types are unknown a priori and evolve dynamically in time. Unlike the traditional bandit setting, observed rewards are generated by a single Markov process. We demonstrate via an illustrative example that blindly applying the traditional bandit algorithms results in very poor performance as measured by regret. We introduce two variants of classical bandit algorithms, upper confidence bound (UCB) and epsilon-greedy, for which we provide theoretical bounds on the regret. We conduct a number of simulation-based experiments to show how the algorithms perform in comparison to traditional UCB and epsilon-greedy algorithms as well as reinforcement learning (Q-learning).