The Robust Design of Experiments: A Review.
Abstract
A review of several optimality criteria for the design of experiments involving quantitative factors for the estimation of unknown parameters in regression problems is given. These include rotatability, D-optimality and G-optimality criteria. Designs are no longer optimal under these criteria when conditions are suboptimal. The two main lines of research in the robust design of experiments will be discussed, i.e., the comparison of the performance of an optimal design with respect to a particular criterion under other criteria, and the construction of designs which guard against particular short comings. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1981
- Accession Number
- ADA100561
Entities
People
- Agnes M. Herzberg
Organizations
- University of Wisconsin–Madison