A graphical diagnostic for heavy tailed data
John P. Nolan · Applied Stochastic Models in Business and Industry · 2020
Abstract Graphical diagnostics are described for general heavy tailed data. This tool allows for a model free assessment of the tails of a univariate dataset a transform on the tails of the data. In addition, one can add to the basic plots comparisons of a dataset to multiple models. Multivariate extensions are described using an ordering based on distance from a center.