Dr. Brian Dandurand is a postdoctoral research fellow working under the supervision of Prof. Andrew Eberhard of RMIT University and Prof. Natashia Boland of the Georgia Institute of Technology. The research project is funded by the Australian Research Council (ARC) grant ARC DP140100985, and the project is titled: Decomposition and Duality: New Approaches to Integer and Stochastic Integer Programming.
His primary research interest is in the design, analysis, and implementation of optimisation algorithms for parallel and/or distributed computing environments. Motivated by this broad interest, more specialised research interests include the nonlinear block Gauss-Seidel methods such as the block coordinate descent method, augmented Lagrangian methods such as the method of multipliers, the alternating direction method of multipliers, proximal bundle methods, and methods related to the Frank-Wolfe method. Recent applications include the development, analysis, and implementation of dual methods in stochastic mixed-integer optimisation and decentralised solution approaches to nonconvex multiobjective optimisation problems, as can be sampled in the selected publications below.
- N.L. Boland, J. Christiansen, B. Dandurand, A. Eberhard, J. Linderoth, J. Luedtke, and F. Oliveira. “Progressive hedging with a Frank-Wolfe based method for computing stochastic mixed-integer programming Lagrangian dual bounds.” 2016.
- N.L. Boland, I. Bakir, B. Dandurand, and A. Erera. “Scenario set partition dual bounds for multistage stochastic programming: A hierarchy of bounds and a partition sampling approach.” 2016.
- B. Dandurand and M.M. Wiecek. “Distributed computation of Pareto sets.” SIAM Journal on Optimization, 25(2):1083--1109, 2015.