Characterizing Uncertainty in Correlated Response Variables for Pareto Front Optimization
Abstract
Current research provides a method to incorporate uncertainty into Pareto front optimization by simulating additional response surface model parameters according to a Multivariate Normal Distribution (MVN). This research shows thatanalogous to the univariate case, the MVN understates uncertainty, leading to overconfident conclusions when variance is not known and there are few observations (less than 25-30 per response). This research builds upon current methods using simulated response surface model parameters that are distributed according to an Multivariate t-Distribution (MVT), which can be shown to produce a more accurate inference when variance is not known. The MVT betteraddresses uncertainty in the parameters which can affect the frequency of treatments appearing on the Pareto front resulting in potentially different proposed solution spaces from that of the MVN.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 26, 2020
- Accession Number
- AD1102509
Entities
People
- Peter A. Calhoun
Organizations
- Air Force Institute of Technology