Skip to main content
Figure 5 | BMC Bioinformatics

Figure 5

From: Classification of bioinformatics workflows using weighted versions of partitioning and hierarchical clustering algorithms

Figure 5

Combined classification results for the Armadillo and my Experiment datasets obtained using the four encoding types with and without weights (average Rand index ± SEM). The unweighted encoding strategies were respectively denoted as Unw I,II (combined results for unweighted encodings of Types I and II) and Unw III,IV (combined results for unweighted encodings of Types III and IV). Panel (a) illustrates the effect of the optimization criteria for both unweighted (first two sets of bars) and weighted (last four sets of bars) encodings; panel (b) - the effect of the distance measure; panel (c) - the effect of the applied partitioning algorithm.

Back to article page