View in EDS

Efficiency of observed information adaptive designs in linear models.

Saved in:
Bibliographic Details
Title: Efficiency of observed information adaptive designs in linear models.
Authors: Lane, Adam1 (AUTHOR) adam.lane@cchmc.org
Source: Journal of Statistical Planning & Inference. Mar2023, Vol. 223, p63-74. 12p.
Subjects: Information design, Concave functions, Fisher information, Sample size (Statistics), Work design, Experimental design
Abstract: The optimization of likelihood based inference through efficient experimental design is considered. A traditional framework for efficiency optimization is to maximize a concave function of expected Fisher information, commonly referred to as optimal design. There exists a substantial amount of literature that suggest that the observed Fisher information is a more relevant measure of the precision of the maximum likelihood estimate. Despite evidence in its favor limited effort has been made to find designs that are optimal with respect to this measure. In this work an adaptive design that incorporates observed Fisher information is proposed. The proposed design is shown to be more efficient than the optimal design with respect to inference. This theoretical result is used to develop a sample size calculation. It is found that fewer samples are required for the proposed design to obtain an equivalent level of precision as the optimal design. • Derivation of observed information adaptive designs for linear models. • Theoretical advantages associated with adaptively incorporating observed information. • Sample size calculation for fixed designs and observed information adaptive designs. [ABSTRACT FROM AUTHOR]
Copyright of Journal of Statistical Planning & Inference is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:The optimization of likelihood based inference through efficient experimental design is considered. A traditional framework for efficiency optimization is to maximize a concave function of expected Fisher information, commonly referred to as optimal design. There exists a substantial amount of literature that suggest that the observed Fisher information is a more relevant measure of the precision of the maximum likelihood estimate. Despite evidence in its favor limited effort has been made to find designs that are optimal with respect to this measure. In this work an adaptive design that incorporates observed Fisher information is proposed. The proposed design is shown to be more efficient than the optimal design with respect to inference. This theoretical result is used to develop a sample size calculation. It is found that fewer samples are required for the proposed design to obtain an equivalent level of precision as the optimal design. • Derivation of observed information adaptive designs for linear models. • Theoretical advantages associated with adaptively incorporating observed information. • Sample size calculation for fixed designs and observed information adaptive designs. [ABSTRACT FROM AUTHOR]
ISSN:03783758
DOI:10.1016/j.jspi.2022.09.001