Quinino, Roberto da Costa (Universidade Federal de Minas Gerais. ICEx. Departamento de Estatística);

Suyama, Emílio (Universidade Federal de Minas Gerais. ICEx. Departamento de Estatística).

Pesquisa Operacional; volume 22, número 1, páginas 1-8. Junho 2002.

control of attributes;

misclassification;

total expected cost.

In this paper we study the optimal number of independent classifications on each produced unit when there are two types of classification errors: the type I error where a product is classified as non-conform when actually it is conform, and the type II error where a product is classified as conform when actually it isn't. We develop an economic model to minimize the total expected cost as a function of the classification errors and the cost of each classification.