Program for Classification of Random Data by Quantile Skewness and Kurtosis

Data classification has proved itself useful both in scientific disciplines and in business. There already exist several classification methods that utilize skewness and kurtosis, particularly for classification of probability distributions, but they rely on the moment-based version of skewness and kurtosis that possess several drawbacks . This project focuses on quantile skewness and kurtosis which have already been used for distribution classification and at the same time describes a newly developed method for parameter identification. We provide a Python module that, based on a sample received from a file or directly from code, produces a list of potential distributions and their parameters. This paper contains 39 pages, 3 chapters, 20 illustrations, 10 bibliography items, 4 appendices. Keywords – data classification, quantile skewness, quantile kurtosis, Pearson System, Python.