Annual Progress Seminar
Monday, 30th September 2023, 11:30 am-12.00 pm
Venue: Room 216, Department of Mathematics
Speaker: Niphadkar Shubham Sanjay
Title: Optimality of Multivariate Crossover Design in Correlated Setup
Abstract: We consider multivariate crossover designs with correlation
between responses. In order to account for correlation between responses,
we consider the proportional and the generalized Markov-type of covariance
structures. We investigate efficiency using various structures for the
variance-covariance matrix within the response, namely the exponential,
gaussian and Mattern(1.5) structures. We prove that an orthogonal array
design of type I and strength 2 is A-, D- and E-optimal for the direct
effects under the proportional covariance structure. We propose the mixed
effects model using subject effects as random effects and obtain
the structure of the information matrix for the direct effects.