N. Bora Keskin

Past Awards

2024
MSOM Young Scholar Prize: Awardee(s)
Best Student Paper Award: Winner(s)
Winning material: Information Provision with Information Overload
MSOM Young Scholar Prize: Winner(s)


2022
Best Paper Competition: First Place
Winning material: Data-driven Clustering and Feature-based Retail Electricity Pricing with Smart Meters


2020
SAS Data Mining Best Paper Award: Winner(s)
Winning material: Data-driven Clustering and Feature-based Retail Electricity Pricing with Smart Meters


2019
SAS Data Mining Best Paper Award: Finalist
Winning material: Personalized Dynamic Pricing with Machine Learning: High Dimensional Features and Heterogeneous Elasticity
Frederick W. Lanchester Prize: Winner(s)
2019 - Winner(s)
Citation:

The Lanchester Prize for 2019 is awarded to Omar Besbes, Yonatan Gur, N. Bora Keskin, and Assaf Zeevi for their series of papers:

  • O. Besbes, Y. Gur, and A. Zeevi, “Non-stationary stochastic optimization,” Operations Research 63, pp. 1227–1244 (September–October 2015)
  • N. B. Keskin and A. Zeevi, “Chasing demand: Learning and earning in a changing environment, ”Mathematics of Operations Research 42, pp. 277–307 (May 2016)
  • O. Besbes, Y. Gur, and A. Zeevi, “Optimal exploration-exploitation in a multi-armed-bandit problem with non-stationary rewards,” Stochastic Systems (Forthcoming)

This set of papers presents the development of a novel paradigm for the modeling and analysis of online dynamic optimization problems in nonstationary environments. Online dynamic optimization, including the multiarmed bandit problem, is a fundamental problem that is prevalent in a wide and disparate range of application domains, such as revenue management, online advertising, clinical trials, and portfolio optimization. To capture nonstationarity in these problems, the authors introduce a “variation budget” that establishes an explicit limit on how much the system parameters can change over time. This modeling innovation allows the authors to present the first complete analytical treatment of this class of problems under nonstationarity. The research results include near-optimal dynamic policies for exploration–exploitation (or learning and earning) and provide a quantification of the “price of nonstationarity.” In one instance, the research develops and characterizes near-optimal pricing policies for a classic revenue management problem in which the demand function can change over time. To date, these papers have been very influential in both the operations research and machine learning communities, and they have inspired important streams of follow-up research.

The Committee members (Stephen Graves, chair, Shane Henderson, Eva Lee, Candace Yano, and Yinyu Ye) are pleased to designate Omar Besbes, Yonatan Gur, N. Bora Keskin, and Assaf Zeevi as recipients of the 2019 Lanchester Prize.



2018
Junior Faculty Forum Paper Competition: Honorable Mention
Winning material: Dynamic Pricing with High Dimensional Covariates and Heterogeneous Elasticity