INFORMS News: Uriel G. Rothblum wins Expository Writing Award

The 2012 INFORMS Expository Writing Award was awarded posthumously to Uriel G. Rothblum, a professor at the Technion-Israel Institute of Technology at the time of his untimely death earlier this year. The award honors an operations researcher whose publications, over a period of at least 10 years, demonstrate a consistently high standard of expository writing.

In announcing the award during the INFORMS Awards Ceremony held in conjunction with the Institute’s annual meeting in Phoenix, Ariz., Prize Committee Richard Steinberg noted Professor Rothblum authored or co-authored more than 160 journal articles addressing such diverse theoretical areas as linear algebra, optimization, dynamic programming, networks, economics, game theory and applied probability, as well as applications to problems found in homeland security and the management of competitive research and development. Professor Rothblum also co-authored (with Frank Hwang) the book “Partitions: Optimality and Clustering” (World Scientific, 2012).

Guy Rothblum’s son accepted the award on his late father’s behalf.

Following are excerpts from the citation:

Professor Rothblum’s most-cited paper is “Algebraic Eigenspaces of Nonnegative Matrices” (Linear Algebra and its Applications, 1975). This paper generalizes the famous Perron-Frobenius theorem regarding the dominant eigenvalue of irreducible nonnegative square matrices to arbitrary nonnegative square matrices. While deriving results in a paper this technical is difficult enough, communicating them clearly requires the careful, concise and precise writing for which Professor Rothblum is known.

An insightful marriage of theory and application is Professor Rothblum’s beautifully written paper with Alvin Roth entitled “Truncation Strategies in Matching Markets – In Search of Advice for Participants” (Econometrica, 1999). Models of matching markets such as those that pair medical residents with hospitals suggest that some residents could improve their individual outcomes by misrepresenting their true preferences. The curious result of this article is that when participants know little about the preferences of others, a good strategy to employ is truncation – that is, restrict the number of positions reported to something lower than the true number acceptable, but truthfully rank those positions that are reported. This paper is structured by moving from a description of matching markets, to a basic model, to a series of cleverly constructed “toy problems”, and finally to several clearly stated theorems together with a summary discussion. In this way, the paper successfully makes its points while heightening reader interest along the way.

Professor Rothblum was also interested in the relationship between optimal solutions in centralized resource allocation problems, and the corresponding equilibria in decentralized versions of the same resource allocation problems. Through the clever use of a system of linear penalties and rewards, Professor Rothblum developed a method by which one could create a game whose equilibrium solution would correspond with the desired centralized optimal solution and thus coordinate the players in the game. This approach was detailed in his paper with Boaz Golany, “Inducing Coordination in Supply Chains through Linear Reward Schemes” (Naval Research Logistics, 2006). The idea is illustrated via a series of clearly and crisply written examples.

An innovative application of this idea in a dynamic setting appears in “A Generalized Two-Agent Location Problem: Asymmetric Dynamics and Coordination” (Journal of Optimization Theory and Applications, 2011, with Golany and Konstantin Kogan). This paper shows how two patrolling agents can coordinate their movement over time to achieve some objective without requiring a centralized controller that yields equivalent results to those that could be achieved via centralized control. The results qualify as superb while the paper itself unfolds like an adventure story.

From his earliest contributions until his untimely death in March of this year, Professor Rothblum tackled very difficult problems in both theoretical and applied mathematics and operations research, yet his writing was always patient and clear. His work exemplifies the art of technical writing, and has been very influential in both theory and applications.