Figure 7From: Using hierarchical clustering methods to classify motor activities of COPD patients from wearable sensor dataClassifier results for Subj C. Dendrogram, results of the LDA, and scatter plot of 1st and 2nd principal components are shown for Subj C. Contrary to what seen for Subj A and B, for Subj C the overlap among clusters is minimal and thus merging clusters does not appear to be necessary.Back to article page