Universal optimality of nearest  neighbour balanced block designs using ARMA models

C. Santharam; K. N. Ponnuswamy

doi:10.6092/issn.1973-2201/978

Universal optimality of nearest neighbour balanced block designs using ARMA models

Authors

C. Santharam Loyola College, Madras
K. N. Ponnuswamy University of Madras

DOI:

https://doi.org/10.6092/issn.1973-2201/978

Abstract

Block designs for observations correlated in one dimension are investigated. Nearest neighbour least balanced designs turnout to be optimal for auto-regressive models when generalized least square estimation is used. Performance of these designs for moving average and ARMA correlation model is shown to be quite satisfactory.

How to Cite

Santharam, C., & Ponnuswamy, K. N. (1995). Universal optimality of nearest neighbour balanced block designs using ARMA models. Statistica, 55(2), 229–238. https://doi.org/10.6092/issn.1973-2201/978