Binay Bhattacharya (SFU)
22/11/13 12:30 Talks
Min-max regret versions of facility location optimization problems
Speaker:
Binay Bhattacharya (Simon Fraser University, Computing Science)
Abstract:
The two most widely considered measures for optimization under uncertainty are minimizing expected cost and minimizing worst-case cost or regret. In stochastic optimization, the parameters of the objective function are governed by a probability distributions, and objective is to find a solution that minimizes the cost. In robust optimization, probabilities are not known, and uncertain parameters are specified either by discrete scenarios or by interval ranges; the objective is to minimize the worst-case cost or regret. In this talk we focus on min-max regret versions with interval structure of uncertainty. We start with a short review of complexity results for the min-max or min-max regret versions of some combinatorial problems. We then discuss a few algorithms to solve center and median problems in restricted networks with uncertain parameters.
This is a joint work with Tiko Kameda and Zhao Song.
Date:
Friday, November 22, 2013
Time:
12:30 pm - 1:30 pm
Location:
TASC1 Building, Room No. 9204, Simon Fraser University, Burnaby
Binay Bhattacharya (Simon Fraser University, Computing Science)
Abstract:
The two most widely considered measures for optimization under uncertainty are minimizing expected cost and minimizing worst-case cost or regret. In stochastic optimization, the parameters of the objective function are governed by a probability distributions, and objective is to find a solution that minimizes the cost. In robust optimization, probabilities are not known, and uncertain parameters are specified either by discrete scenarios or by interval ranges; the objective is to minimize the worst-case cost or regret. In this talk we focus on min-max regret versions with interval structure of uncertainty. We start with a short review of complexity results for the min-max or min-max regret versions of some combinatorial problems. We then discuss a few algorithms to solve center and median problems in restricted networks with uncertain parameters.
This is a joint work with Tiko Kameda and Zhao Song.
Date:
Friday, November 22, 2013
Time:
12:30 pm - 1:30 pm
Location:
TASC1 Building, Room No. 9204, Simon Fraser University, Burnaby