Clustering gene expression data using a diffraction‐inspired framework

Table 4 Comparison of the clustering results for the Shipp dataset

Algorithm	AEC	AVI	ACC (%)
Diffractive clustering	55.5	160.6	63.6
k‐means	48.0	79.7	51.9
Fuzzy c‐means	48.7	82.0	51.9
Hierarchical clustering (single)	51.1	112.2	53.3
Hierarchical clustering (centroid)	51.1	113.6	53.2
Self‐organising map	47.7	14.7	53.3
Gaussian Mixture Model	49.8	66.8	51.9
DBSCAN	49.9	65.4	54.5

The diffractive clustering algorithm outperforms the others in terms of accuracy, however all the algorithms perform relatively poorly due to the mixing between the two different classes.

ISSN: 1475-925X