Past Awards
The book has arguably the most comprehensive treatment of algorithmic continuous optimization. By developing both foundational results and recent breakthroughs, its publication has advanced the state of the art in operations research and its extensions in science, engineering, computer science, and machine learning.
Yurii Nesterov is the world’s leading authority on the efficiency of algorithms for continuous optimization. His text, Interior-Point Polynomial Algorithms for Convex Programming, co-authored with A. Nemirovskii, utilized the theory of self-concordant functions to unify global complexity results obtained for convex optimization problems including linear, second-order cone and semidefinite programming. A subsequent paper co-authored with M. Todd introduced the theory of self-scaled cones to unify the theory of primal-dual algorithms for these same problem classes. Nesterov is also the author of the text Introductory Lectures on Convex Optimization, which develops state-of-the-art theory at a level appropriate for introductory graduate courses. In recent work he has obtained improved results on the global convergence of a regularized Newton’s method for unconstrained optimization and established a theory of smoothing that allows for the applicability of optimal first-order methods to large-scale problems with nondifferentiable objectives. In 2000, Nesterov received the Dantzig Prize, jointly awarded by the Mathematical Programming Society and the Society for Industrial and Applied Mathematics, in recognition of his contributions to the theory of convex optimization.
Yinyu Ye has been at the forefront of research on interior-point methods and applications of conic optimization for over 20 years. Ye was one of the first researchers to investigate the theory of interior-point methods following the announcement of Karmarkar’s algorithm in 1984 and is responsible for many fundamental results, including the first potential reduction algorithm with a complexity of O(n3 L) operations, the primal-dual predictor-corrector framework (with S. Mizuno and M. Todd) and an interior-point algorithm for linear programming whose complexity is independent of the problem’s objective coefficients and right-hand-side vector (with S. Vavasis). His text, Interior-Point Algorithms: Theory and Analysis, provided the first comprehensive treatment of the topic at a level accessible to non-specialists. He is also the co-author (with D. Luenberger) of the third edition of the well-known text Linear and Nonlinear Programming. In recent years he has developed theory and computational methodology for sensor network location based on semidefinite programming and established new complexity results for problems concerning the computation of economic equilibria. A prolific researcher, Ye is well known for his enthusiasm and generosity in sharing research topics with colleagues and students.