【Academic Seminar】Ambiguous Chance-constrained Binary Programs under Mean-covariance Information

Topic: Ambiguous Chance-constrained Binary Programs under Mean-covariance Information

Speaker:Yiling Zhang, University of Michigan (UM), Ann Arbor

Time and Date: 2:00 pm – 3:00 pm, December 24, 2018

Venue: Boardroom, Dao Yuan Building

Abstract：

We consider chance-constrained binary programs that have wide applications in cloud computing, surgery planning, and many other stochastic resource allocation problems. With available mean and covariance information, we solve distributionally robust chance-constrained binary programs (DCBP). Using two different ambiguity sets, we equivalently reformulate the DCBPs as 0-1 second-order cone (SOC) programs. We further utilize the submodularity of 0-1 SOC constraints and lifting to derive extended polymatroid inequalities. We incorporate the valid inequalities in a branch-and-cut algorithm for efficiently solving DCBPs. We conduct computational studies and demonstrate the performance of DCBP solutions using surgery block allocation instances.

Biography:

Yiling Zhang is a Ph.D. candidate in the Department of Industrial and Operations Engineering (IOE) at the University of Michigan (UM), Ann Arbor. She received her bachelor degree in Automation from Tsinghua University in China in 2013. Her research mainly focuses on the theories of distributionally robust optimization (DRO), utilizing stochastic programming, integer programming, and nonlinear optimization techniques for deriving reformulations and efficient computational methods. Applications of her research include energy, healthcare, and transportation. She is a student fellow of Michigan Institute for Computational Discovery and Engineering (MICDE) and was nominated for the UM College of Engineering Richard & Eleanor Towner Prize for Outstanding Ph.D. Research. She has received the IOE Murty Prize for Best Student Paper in Optimization in 2018 and a Rackham Graduate School International Student Fellowship in 2014.