Optimal Regression Designs Under the Second-Order Least Squares Estimator

• Activity Number: 414 - Advances in Estimation Methods
• Type: Contributed
• Date/Time: Tuesday, July 31, 2018 : 2:00 PM to 3:50 PM
• Sponsor: SSC
• Author(s): Julie Zhou* and Chi-Kuang Yeh
• Companies: University of Victoria and University of Victoria
• Keywords: optimal design; support point; equivalence theorem; nonlinear scale invariance; convex optimization

Abstract

We investigate optimal designs under the second-order least squares estimator for linear and nonlinear regression models. We derive the number of support points in A-, c- and D-optimal designs for various models analytically, and we also discuss numerical algorithms for finding optimal designs for any linear and nonlinear regression models. The algorithms are based on the semi-definite programming in convex optimization, and they are powerful to solve optimal design problems with hundreds of variables. Several applications will be presented.