Cluster validation using the non-parametric bootstrap and parallel processing: applications in unsupervised machine learning of Shimodaira's method to text mining and genomics

Abstract

Cluster analysis is a type of machine learning used in many areas of research. Cluster validation is a method of determining a level of confidence for the results of cluster analysis. The goal of this research was to write a program trueTree that would perform cluster validation. trueTree proved successful. Every time trueTree's results were compared to past research, the results matched confirming that trueTree works properly.