Title | sigclust2: Statistical Significance for Hierarchical Clustering (R) |
Publication Type | Software |
Year of Publication | 2016 |
Authors | Kimes, Patrick |
Version | 1.2.4 |
Type | R |
Abstract | This package may be used to assess statistical significance in hierarchical clustering. To assess significance in high-dimensional data, the approach assumes that a cluster may be well approximated by a single Gaussian (normal) distribution. Given the results of hierarchical clustering, the approach sequentially tests from the root node whether the data at each split/join correspond to one or more Gaussian distributions. The hypothesis test performed at each node is based on a Monte Carlo simulation procedure, and the family-wise error rate (FWER) is controlled across the dendrogram using a sequential testing procedure. |
Project:
Software Weblinks: