|
Activity Number:
|
396
- Distributional Robustness, Validity, Causality, and Generalizability
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 12, 2021 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
IMS
|
|
Abstract #316968
|
|
|
Title:
|
Remedying Estimation Unobservability via Distributionally Robust Optimization
|
|
Author(s):
|
Henry Lam*
|
|
Companies:
|
Columbia
|
|
Keywords:
|
distributionally robust optimization;
simulation;
extreme event analysis;
uncertainty quantification
|
|
Abstract:
|
Distributionally robust optimization (DRO), a surging methodology for optimization under uncertainty, advocates worst-case analyses over unknowns within so-called ambiguity sets that capture information or modelers' beliefs. We explain how this rationale can be used to construct statistically justified estimators in extreme event analysis and simulation uncertainty quantification where relevant data could be unobservable. We discuss some tradeoffs concerning conservativeness, and also the underpinning probabilistic and optimization tools in tackling the DRO problems arising from these estimation goals.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2021 program
|
