Chance-constrained optimization under limited distributional information: A review of reformulations based on sampling and distributional robustness

Chance-constrained programming (CCP) is one of the most difficult classes optimization problems that has attracted attention researchers since 1950s. In this survey, we focus on cases when only limited information distribution available, such as a sample from distribution, or moments distribution. We first review recent developments in mixed-integer linear formulations chance-constrained programs arise finite discrete distributions (or average approximation). highlight successful reformulations and decomposition techniques enable solution large-scale instances. then active research distributionally robust CCP, which framework to address ambiguity random data. The focal point our scalable can be readily implemented with state-of-the-art software. Furthermore, prevalence CCPs applications across multiple domains.