Optimal design strategies for relative potency using the two-parameter log-logistic model

Article type: Research Article

Authors: Tusto, Pattaraporn^{a; c} | O'Brien, Timothy E.^b | Tiensuwan, Montip^{a; c; *}

Affiliations: [a] Department of Mathematics, Faculty of Science, Mahidol University, Bangkok, Thailand | [b] Department of Mathematics and Statistics, Loyola University, Chicago, IL, USA | [c] Center of Excellence in Mathematics, CHE, Bangkok, Thailand

Correspondence: [*] Corresponding author: Montip Tiensuwan, Department of Mathematics, Faculty of Science, Mahidol University, Bangkok, Thailand. E-mail:montip.tie@mahidol.ac.th

Abstract: In this paper, we focus on the D- and Ds-optimal designs for two-parameter log-logistic (LL2) relative potency model where the response variables are normal and binomial distributions. The D- and Ds-optimal designs are obtained by using D-optimal design and nesting strategy criterions, respectively. Furthermore, the general equivalence theorems is used to guarantee the D- and Ds-optimal designs. The results show that we obtain four support points for D-optimal designs and two support points for Ds-optimal designs.

Keywords: General equivalence theorem, optimal design, relative potency, two-parameter log-logistic model

DOI: 10.3233/MAS-150357

Journal: Model Assisted Statistics and Applications, vol. 11, no. 2, pp. 109-123, 2016