INFORMS NEWS: Kelly wins Saul Gass Expository Writing Award

Frank P. Kelly, professor of Mathematics of Systems at the University of Cambridge University of Cambridge, was named the 2013 recipient of the Saul Gass Expository Writing Prize at the INFORMS Annual Meeting in Minneapolis. Committee Chair Edward Kaplan accepted the award on Kelly’s behalf.

Recently named in honor of Saul Gass, an O.R. pioneer and extraordinary and prolific writer who passed away in March, the prize recognizes an operations researcher/management scientist whose publications demonstrate a consistently high standard of expository writing.

The citation reads in part:

A Fellow of Britain’s Royal Society and Commander of the Order of the British Empire, Professor Kelly’s seminal research – in the theory of stochastic networks and applications of this theory to the design and control of communications systems – has been recognized with many prior awards, including the INFORMS Lanchester and von Neumann prizes. The present award celebrates the excellence of Professor Kelly’s expository writing.

Professor Kelly’s book “Reversibility and Stochastic Networks,” first published in 1979 and subsequently reprinted in 1987, 1994 and 2011, has been cited more than 2,500 times. In this work, Professor Kelly succinctly and elegantly developed the concepts of reversibility and quasi-reversibility, and explained how they simplify the analysis of many stochastic networks.

While the exposition is rigorous, even the most advanced sections are accessible to readers with only a basic knowledge of probability theory. Also of note are the many thought-provoking exercises at the end of each section, for writing good exercises is hard. One published review states that, “The book is truly well written … elegant and intellectually stimulating without being unnecessarily general or abstract … In clarity, generality, and elegance there is no close competitor” [1], while a second reviewer reports that, “Part of the reason the mathematics is so clear is that Kelly’s pleasant, simple English exploits fully the power of natural language to describe abstract processes” [2].

Those who read this book are unlikely to forget Professor Kelly’s initial description of reversibility: “Speaking intuitively, if we take a film of such a process and then run the film backwards the resulting process will be statistically indistinguishable from the original process.”

Professor Kelly has also authored or coauthored roughly 100 articles that contribute to the literature in many different areas, including stochastic networks, queueing theory, communication networks, population biology, search theory, transportation and group decision-making. To highlight one example, Professor Kelly’s most-cited paper, “Rate Control for Communication Networks: Shadow Prices, Proportional Fairness and Stability” (co-authored with A.K. Maulloo and D.K.H. Tan, Journal of the Operational Research Society, 1998), has been referenced nearly 4,250 times.

Building upon his earlier work, this article joins stability, an engineering issue, with fairness, an economic issue, in developing tractable models for allocating flow rates across the users of a communication network. The authors provide decentralized schemes by which a network can manage itself, and the resulting bandwidth allocations have the property that the amount a user is willing to pay is equal to the amount charged by the network. As appropriate for an article modeling communications traffic, the writing “flows” through the paper’s structural “nodes” of problem description, model formulation, theoretical analysis and numerical examples. It is an exciting paper to read.

Professor Kelly’s exposition is careful, patient and clear. His work exemplifies the art of technical writing, and has been extremely influential in both theory and applications.

Robert A. Shumsky and Dimitris Bertsimas joined Kaplan on the prize committee.

References

  1. Networks, Vol. 13, No. 1, pp. 153-154; 1983.
  2. Technometrics, Vol. 23, No. 2, pp. 206-207; 1981.