Robust optimization
Prof. Aharan Ben-Tal, Professor of Operations Research at the Faculty of Industrial Engineering and Management at the Technion – Israel Institute of Technology, will give the following two seminars for the ABRO PhD course:
- March 13 (10:00-12:00) Room A4 Robust optimization: the need, the challenge, the success.
- March 14 (10:00-12:00) Room Aula Magna Multi-stage (dynamic) optimization problems affected by uncertainty
ABSTRACT
LECTURE 1
We list and illustrate by examples various sources of uncertainty associated with optimization problems. We then explain the difficulties arising when solving such uncertainty affected problems due to lack of full information on the nature of the uncertainty on one hand, and the likelihood of facing computationally intractable problems on the other hand. Robust Optimization (RO) is a methodology that was designed from the start to meet the above challenges. We will review the theory underlying the RO methodology and demonstrate its success in solving meaningful static conic problems (linear, conic quadratic and semidefinite programs) affected by uncertainty, and demonstrate their success in solving some challenging Engineering problems.
LECTURE 2
We introduce RO methodology, based on (mainly) linear decision rules, to treat multi- stage optimization problems affected by uncertainty. The methodology is used to solve a supply chain problem.We then address problems, including Chance Constrained ones, where stochastic parameters suffer from distributional ambiguity, and show how RO can provide tractable approximations for these hard problems.
