Data-Driven Stochastic Robust Optimization: A General Computational Framework and Algorithm for Optimization under Uncertainty in the Big Data Era

07/28/2017 ∙ by Chao Ning, et al. ∙ 0

A novel data-driven stochastic robust optimization (DDSRO) framework is proposed to systematically and automatically handle labeled multi-class uncertainty data in optimization problems. Uncertainty realizations in large datasets are often collected from various conditions, which are encoded by class labels. A group of Dirichlet process mixture models is employed for uncertainty modeling from the multi-class uncertainty data. The proposed data-driven nonparametric uncertainty model could automatically adjust its complexity based on the data structure and complexity, thus accurately capturing the uncertainty information. A DDSRO framework is further proposed based on the data-driven uncertainty model through a bi-level optimization structure. The outer optimization problem follows a two-stage stochastic programming approach to optimize the expected objective across different classes of data; robust optimization is nested as the inner problem to ensure the robustness of the solution while maintaining computational tractability. A tailored column-and-constraint generation algorithm is further developed to solve the resulting multi-level optimization problem efficiently. Case studies on strategic planning of process networks are presented to demonstrate the applicability of the proposed framework.

READ FULL TEXT
POST COMMENT

Comments

There are no comments yet.

Authors

page 1

page 2

page 3

page 4

This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.