John von Neumann Theory Prize

The 2012 John von Neumann Theory Prize of INFORMS is awarded to George Nemhauser and Laurence Wolsey for their outstanding and lasting contributions to integer optimization and example setting scholarship.  Both individually and jointly, they have advanced significantly our understanding of discrete optimization both from theoretical and practical perspectives.

George Nemhauser is one of the most influential scholars in the operations research and optimization community. In his over 50 years career (his Phd was in 1961), George has advanced the theory and practice of discrete optimization through numerous articles, influential books and computer codes. He is the only researcher who has won the Lanchester prize twice. In 1978, he (together with G. Cornuejols and M. Fisher) won the Lanchester prize for his pioneering analysis of an approximation algorithm for a facility location problem. In 1989, George and Laurence won the Lanchester prize for their book "Integer Programming and Combinatorial Optimization’’.  He is a member of their National Academy of Engineering and has received the Kimball medal.

Laurence Wolsey has a similarly long career and has become one of the most recognized members of our community due to his sustained and constant high-level contributions to many aspects of optimization. Earlier this summer Laurence was the recipient of the Dantzig prize for contributing significantly to foundational understanding of the geometry of mixed-integer optimization, to duality theory in discrete optimization, and to the development of effective new methods to the variety of applications, particularly in production planning and scheduling.

George and Laurence pioneered the study of polyhedral combinatorics when it was not yet fashionable and practically successful to use cutting planes and the like in integer programming. They both contributed to facility location, cutting stock and stochastic programming, and in particular to various aspects of production and production planning.

George and Laurence have jointly published more than 10 papers that range from a recursive procedure to generate all cuts for 0-1-mixed-integer program, via uncapacitated facility location, to maximization of submodular set functions and worst-case and probabilistic analysis of algorithms as well as travelling salesman problems. Their   joint book "Integer and Combinatorial Optimization" has had a significant influence on the community. Their development of mixed-integer rounding (MIR)   are the prime tools nowadays in general codes for the solution of integer and mixed-integer programming problems.

They also contributed significantly to the development of codes for the solution of general or specific integer programming problems. The MINTO development that George did (together with Savelsbergh and Sigismondi) was a precursor of modern branch-and-cut codes.  The book "Production Planning by Mixed-integer Programming" by Laurence and Yves Pochet is a pioneering extensive monograph showing how to model and solve relevant planning problems.

Both George and Laurence have been outstanding research supervisors and have influenced the research directions of many other younger colleagues by their advice and guidance. They both share their insights and graciously provide ideas to many other colleagues.