Estimation Bootstrap of the Probability Distributions of the Loss Function of Taguchi and of the Signal-To-Noise Ratio
Marcos Augusto Mendes Marques, Anselmo Chaves Neto

Abstract
The knowledge of the probability distributions in experiments it is necessary for the accomplishment of statistical analyses of the experimental data. Such analyses have as objective to obtain relative information to the population of origin of the data supplied by random samples. This way, applied statistical inferences to the quality indicators as the loss function of Taguchi and signal to noise ratio should be used as a method for taking decisions regarding the quality of a product (service) and of the process of manufacture itself. Such indicators, together, measure the bad quality and robustness of a product in relation to variable answer of a functional characteristic. As it is necessary the knowledge of the probability distributions for the accomplishment of those inferences and, as in general, they are ignored is reasonable the application of the method intensive computationally of resample known like Bootstrap, for an estimate of those distributions.

Full Text: PDF