scholarly journals Optimal Transport-Based Distributionally Robust Optimization: Structural Properties and Iterative Schemes

2021 ◽  
Author(s):  
Jose Blanchet ◽  
Karthyek Murthy ◽  
Fan Zhang
Keyword(s):  
Robust Optimization ◽  
Optimal Transport ◽  
Decision Rules ◽  
Stochastic Gradient Descent ◽  
Transport Cost ◽  
Distributionally Robust Optimization ◽  
Worst Case ◽  
Strongly Convex ◽  
Distributionally Robust ◽  
Convexity Assumptions

We consider optimal transport-based distributionally robust optimization (DRO) problems with locally strongly convex transport cost functions and affine decision rules. Under conventional convexity assumptions on the underlying loss function, we obtain structural results about the value function, the optimal policy, and the worst-case optimal transport adversarial model. These results expose a rich structure embedded in the DRO problem (e.g., strong convexity even if the non-DRO problem is not strongly convex, a suitable scaling of the Lagrangian for the DRO constraint, etc., which are crucial for the design of efficient algorithms). As a consequence of these results, one can develop efficient optimization procedures that have the same sample and iteration complexity as a natural non-DRO benchmark algorithm, such as stochastic gradient descent.

scholarly journals The Distributionally Robust Optimization Reformulation for Stochastic Complementarity Problems

2014 ◽  
Vol 2014 ◽  
pp. 1-7 ◽  
Author(s):  
Liyan Xu ◽  
Bo Yu ◽  
Wei Liu
Keyword(s):  
Robust Optimization ◽  
Linear Complementarity Problem ◽  
Complementarity Problem ◽  
Linear Complementarity ◽  
Uncertain Parameters ◽  
Distributionally Robust Optimization ◽  
Worst Case ◽  
Nonnegativity Constraints ◽  
Distributionally Robust ◽  
Stochastic Linear Complementarity Problem

We investigate the stochastic linear complementarity problem affinely affected by the uncertain parameters. Assuming that we have only limited information about the uncertain parameters, such as the first two moments or the first two moments as well as the support of the distribution, we formulate the stochastic linear complementarity problem as a distributionally robust optimization reformation which minimizes the worst case of an expected complementarity measure with nonnegativity constraints and a distributionally robust joint chance constraint representing that the probability of the linear mapping being nonnegative is not less than a given probability level. Applying the cone dual theory and S-procedure, we show that the distributionally robust counterpart of the uncertain complementarity problem can be conservatively approximated by the optimization with bilinear matrix inequalities. Preliminary numerical results show that a solution of our method is desirable.

scholarly journals Performance evaluation for distributionally robust optimization with binary entries

2020 ◽  
Vol 11 (1) ◽  
pp. 1-9
Author(s):  
Shunichi Ohmori ◽  
Kazuho Yoshimoto
Keyword(s):  
Performance Evaluation ◽  
Robust Optimization ◽  
Optimization Technique ◽  
Data Driven ◽  
Optimal Decision ◽  
Distributionally Robust Optimization ◽  
Worst Case ◽  
Leibler Divergence ◽  
Out Of Sample ◽  
Distributionally Robust

We consider the data-driven stochastic programming problem with binary entries where the probability of existence of each entry is not known, instead realization of data is provided. We applied the distributionally robust optimization technique to minimize the worst-case expected cost taken over the ambiguity set based on the Kullback-Leibler divergence. We investigate the out-of-sample performance of the resulting optimal decision and analyze its dependence on the sparsity of the problem.

scholarly journals Partition-based distributionally robust optimization via optimal transport with order cone constraints

4OR ◽  
2021 ◽  
Author(s):  
Adrián Esteban-Pérez ◽  
Juan M. Morales
Keyword(s):  
Robust Optimization ◽  
Transport Theory ◽  
Optimal Transport ◽  
A Priori ◽  
Uncertain Parameters ◽  
Finite Sample ◽  
Distributionally Robust Optimization ◽  
Cone Constraints ◽  
Wide Range ◽  
Distributionally Robust

AbstractIn this paper we wish to tackle stochastic programs affected by ambiguity about the probability law that governs their uncertain parameters. Using optimal transport theory, we construct an ambiguity set that exploits the knowledge about the distribution of the uncertain parameters, which is provided by: (1) sample data and (2) a-priori information on the order among the probabilities that the true data-generating distribution assigns to some regions of its support set. This type of order is enforced by means of order cone constraints and can encode a wide range of information on the shape of the probability distribution of the uncertain parameters such as information related to monotonicity or multi-modality. We seek decisions that are distributionally robust. In a number of practical cases, the resulting distributionally robust optimization (DRO) problem can be reformulated as a finite convex problem where the a-priori information translates into linear constraints. In addition, our method inherits the finite-sample performance guarantees of the Wasserstein-metric-based DRO approach proposed by Mohajerin Esfahani and Kuhn (Math Program 171(1–2):115–166. 10.1007/s10107-017-1172-1, 2018), while generalizing this and other popular DRO approaches. Finally, we have designed numerical experiments to analyze the performance of our approach with the newsvendor problem and the problem of a strategic firm competing à la Cournot in a market.

STOCHASTIC MODEL PREDICTIVE CONTROL FOR SPACECRAFT RENDEZVOUS AND DOCKING VIA A DISTRIBUTIONALLY ROBUST OPTIMIZATION APPROACH

2021 ◽  
pp. 1-19
Author(s):  
ZUOXUN LI ◽  
KAI ZHANG
Keyword(s):  
Model Predictive Control ◽  
Stochastic Model ◽  
Robust Optimization ◽  
Predictive Control ◽  
Attitude Control ◽  
Optimization Approach ◽  
Distributionally Robust Optimization ◽  
Distributionally Robust ◽  
Stochastic Model Predictive Control ◽  
Rendezvous And Docking

Abstract A stochastic model predictive control (SMPC) algorithm is developed to solve the problem of three-dimensional spacecraft rendezvous and docking with unbounded disturbance. In particular, we only assume that the mean and variance information of the disturbance is available. In other words, the probability density function of the disturbance distribution is not fully known. Obstacle avoidance is considered during the rendezvous phase. Line-of-sight cone, attitude control bandwidth, and thrust direction constraints are considered during the docking phase. A distributionally robust optimization based algorithm is then proposed by reformulating the SMPC problem into a convex optimization problem. Numerical examples show that the proposed method improves the existing model predictive control based strategy and the robust model predictive control based strategy in the presence of disturbance.

Sign in / Sign up

Export Citation Format

Share Document