Classification of subjects in the tuberculin test: application of a statistical model when the frequency distribution of the induration is a mixture of normal components

The classification problem in the tuberculin test was studied. A sample of induration data from a generic population was used and the "best possible regions" criterion was applied. Estimates of classification regions and estimates of probabilities of misclassification were obtained.

Tuberculin test; Induration; Mixture of normal distributions; Classification regions; Misclassification probability