Skip to main content
Figure 4 | BMC Bioinformatics

Figure 4

From: A network-assisted co-clustering algorithm to discover cancer subtypes based on gene expression

Figure 4

Accuracies on simulated datasets. NCIS (α = 0.85) vs. NCIS (α = 0) vs. consensus clustering on simulated datasets. Height of the solid boxes reflects the average accuracy in each setting (over 5 independent datasets simulated under the setting) and the bar indicates the standard deviation. P-value of paired one-sided t-test (25 data points for each group) for H0: Accuracy (NCIS (α = 0.85)) ≤ Accuracy (NCIS (α = 0)) is 0.0057. P-value of paired one-sided t-test (25 data points for each group) for H0: Accuracy (NCIS (α = 0.85)) ≤ Accuracy (Consensus clustering) is 0.0019.

Back to article page