Statistics for Hierarchical Clustering with Multiple-Height Branch-Cut Applied to Short Time-Series Gene Expression Data