THE MINIMUM INACCURACY PRINCIPLE IN ESTIMATING POPULATION PARAMETERS FROM GROUPED DATA

Abstract

In a previous paper the minimum inaccuracy principle was suggested as an operative method for estimating population parameters when the available experimental information could not be perceived as an exact outcome, but rather as fuzzy information. This principle is an extension of the maximum likelihood principle of estimating from exact experimental data. In this paper, the particularization of the first method to the case in which each fuzzy information reduces to a class of extact observations is developed. We then analyze certain correspondence between the maximum likelihood and minimum inaccuracy principles in estimating parameters after grouping data. In addition, we prove that the second method approximates to the first one when a certain natural grouping, or choice of classes, is accomplished. Finally, in order to illustrate the preceding results, some relevant particular cases are examined.

Citation

GIL, M.A. and CORRAL, N. (1987), "THE MINIMUM INACCURACY PRINCIPLE IN ESTIMATING POPULATION PARAMETERS FROM GROUPED DATA", Kybernetes, Vol. 16 No. 1, pp. 43-49. https://doi.org/10.1108/eb005756