Optimal experimental design that minimizes the width of simultaneous confidence bands

Kuriki, Satoshi and Wynn, Henry P. ORCID: 0000-0002-6448-1080 (2019) Optimal experimental design that minimizes the width of simultaneous confidence bands. Electronic Journal of Statistics, 13 (1). pp. 1099-1134. ISSN 1935-7524

Text (Optimal experimental design that minimizes the width of simultaneous confidence bands) - Accepted Version Download (399kB)

• Author

Abstract

We propose an optimal experimental design for a curvilinear regression model that minimizes the band-width of simultaneous confidence bands. Simultaneous confidence bands for curvilinear regression are constructed by evaluating the volume of a tube about a curve that is defined as a trajectory of a regression basis vector (Naiman, 1986). The proposed criterion is constructed based on the volume of a tube, and the corresponding optimal design that minimizes the volume of tube is referred to as the tube-volume optimal (TV-optimal) design. For Fourier and weighted polynomial regressions, the problem is formalized as one of minimization over the cone of Hankel positive definite matrices, and the criterion to minimize is expressed as an elliptic integral. We show that the Mö bius group keeps our problem invariant, and hence, minimization can be conducted over cross-sections of orbits. We demonstrate that for the weighted polynomial regression and the Fourier regression with three bases, the tube-volume optimal design forms an orbit of the Mö bius group containing D-optimal designs as representative elements.

Item Type:	Article
Additional Information:	© 2019 The Authors
Divisions:	Centre for Analysis of Time Series
Subjects:	Q Science > QA Mathematics
Date Deposited:	28 Jun 2019 13:45
Last Modified:	17 Oct 2024 17:26
URI:	http://eprints.lse.ac.uk/id/eprint/101089

Actions (login required)

View Item