Robust optimization is a distinct approach to optimizations problems that allows for the incorporation of uncertainty. The usefulness of robust optimization lies in the ability to solve for every realization of the uncertain parameters. As a result, the problem can be solved for the worst-case scenarios of the entire set of uncertainty. The most vital aspect of robust optimization is the determination of the uncertainty set. As an uncertainty set grows, it will be able to undertake more realizations. The drawback from assuming a large uncertainty set is the concern of the overall optimization problem becoming computationally intractable. At the same time, a small uncertainty set will yield an answer that is conservative and ignorant of aspects of the uncertainty. The general concepts and usage of robust optimization are now being shifted because of the availability and shear volume of data for every aspect of life. From complex supply chains to internet user preferences, these data is forcing change in how robust optimization problems are being approached. The existence of these data can eliminate the need for unproven assumptions and reasoning, which were previously needed in many robust optimization problems in order to make the problems tractable.

Last modified

• 12/04/2018

Creator

• Watson Fu

DOI

• https://doi.org/10.21985/N2RR19

Keyword

• Data Driven

Rights statement

• In Copyright