Skip to main content
Figure 1 | BMC Bioinformatics

Figure 1

From: Partition decoupling for multi-gene analysis of gene expression profiling data

Figure 1

two circles examples. In (a) and (b), colors denote cluster assignments from k-means (k = 2) and spectral clustering, respectively. In (a), k-means using k = 2 produces a linear cut through the data; in (b), spectral clustering automatically chooses two clusters and assigns clusters with nonconvex boundaries. The spectrally embedded data used in (b) is shown in (c); in this representation, the clusters are linearly separable, and a rug plot shows the bimodal density of the Fiedler vector that yielded the correct number of clusters.

Back to article page