The Identification of the Type of Nonlinear Transformation in Homogeneous Observation Spaces Using the Johnson Transformation.

Abstract

In multivariate data analysis, the assumption of normally distributed observations is crucial, as it significantly simplifies the task of estimating probability functions. However, in practice, marginal distributions can be substantially skewed, necessitating the search for methods to normalize them. This paper presents the results of a study on the effectiveness of the Johnson transformation applied to data obtained through the nonlinear transformation x^α in homogeneous observation spaces.