Showing papers by "Rahul Mukerjee published in 1998"
TL;DR: In this article, the authors make a systematic study of estimation capacity, a criterion of model robustness, under the absence of interactions involving three or more factors, and obtain general results, providing designs with maximum estimation capacity.
Abstract: Using the approach of finite projective geometry, we make a systematic study of estimation capacity, a criterion of model robustness, under the absence of interactions involving three or more factors. Some general results, providing designs with maximum estimation capacity, are obtained. In particular, for two-level factorials, it is seen that constructing a design with maximum estimation capacity calls for choosing points from a finite projective geometry such that the number of lines is maximized and the distribution of these lines among the chosen points is as uniform as possible. We also explore the connection with minimum aberration designs under which the sizes of the alias sets of two-factor interactions which are not aliased with main effects are the most uniform possible.
68 citations
TL;DR: In this article, the authors considered a theoretically interesting class of BOMAs, given by the proper BOMA, and examined their existence and construction problems, and showed that proper bOMAs yield optimal experimental plans and can thus be useful in design of experiments as well.
6 citations
TL;DR: In this paper, a symmetric orthogonal array LN(sn) is defined as an Nxn array with entries from a set of s (~ 2) distinct symbols such that all the s 2 possible ordered pairs of symbols appear equally often as row vectors in every N x 2 subarray thereof.
Abstract: Since X -m and X -m-1 are consecutive integers, the inequality (I) follows considering the expectation of the nonnegative quantity (x-m)(x-m-1). The simple inequaliy \\(1) is useful in proving nonexistence results on symmetric rothogonal arrays. For ease in reference, recall that a symmetric orthogonal array LN(sn), of strenths two, is an Nxn array, with entries from a set of s (~ 2) distinct symbols such that all the s2 possible ordered pairs of symbols appear equally often as row vectors in every N x 2 subarray thereof. In LN(sn), let x; be the number of positions in which the 1st and the ith rows have indentical entries (2 ::; i ::; N). Also, let X be an integer-valued random variable assuming values x2 , ... , XN, each with probaility l/(N -1). Then, from the definition of LN(sn), it is not hard to see that