Models for Transformation: A Global Optimization Transformation method with Some Extension from Box-Cox Transformation

Abstract

The choice of a transformation has often been made in an ad hoc trial-and-error fashion but in this research paper we deals with the new solution of transformation which is covering all the real numbers. The assumption of normality that gained from the optimization method is much better improved compared to Box-Cox transformation through the statistical P value. The biggest values of the statistical P value (> 0.05) reflect the goodness of the normality achievement. In order to obtain the efficiency status, we will illustrate the application of transformation method with the data that are getting from Hospital University Science Malaysia (HUSM).