.png){kind=icon}
Return
Data-Driven Robust Optimization Based on Principle Component Analysis and Cutting Plane Methods
.png){kind=icon}
PRE
AI
DOI：10.1021/acs.iecr.1c03886.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
A data-driven robust optimization method is proposed by leveraging the merits of machine learning for decision-making under uncertainty. Robust optimization tends to result in performance loss due to its conservatism. To hedge against the conservational results, a data-driven general polyhedral uncertainty set is constructed by using the principal component analysis (PCA) and the cutting plane methods. By employing PCA to uncertainty data, correlations and distribution knowledge between uncertain variables are effectively captured. To remove redundant scenarios in the uncertainty set, cutting planes are introduced and an asymmetrical structure is generated. Besides, the statistic limits are utilized to construct a probability uncertainty set to realize the trade-off between the conservatism and protection level. The constructed uncertainty sets are incorporated into data-driven static and two-stage adaptive optimization models to induce robust counterparts. Finally, the efficiency of the proposed framework is verified by using a numerical experiment and the application of batch production scheduling.
Keywords:
DECISION-MAKING
UNCERTAINTY
DESIGN
FRAMEWORK
Journal
I
IF:
3.9
Papers:
4.0W
Citations:
9.6W
Organization
Cited Papers
.png){kind=icon}
ShareSaveShareSave
ShareSave
View more