Skip to main content
Figure 9 | BMC Bioinformatics

Figure 9

From: BicSPAM: flexible biclustering using sequential patterns

Figure 9

Comparing BicSPAM discretization options. The use fixed ranges is the simplest discretization option, but commonly leads to an accentuated weak distribution of items and is prone to the items-boundary problem. Percentage-based method tackle this observation using a depth partitioning of items that leads to intervals containing approximately the same number of items. Finally, alternative distributions (as the illustrated Gaussian) can be adopted to combine the properties of the previous solutions. Although Gaussian distributions are typically selected, Poisson distributions with a considerable number of occurrences (λ≥3) are dynamically selected for datasets without symmetric distribution of values around the median value. As illustrated, these methods can lead to biclustering solutions with heightened differences.

Back to article page