Skip to main content

Table 3 Optimal Cluster configuration by graph type and resilience measures

From: Applications of node-based resilience graph theoretic framework to clustering autism spectrum disorders phenotypes

  Complete clustering No node reassignment
  Integrity k=3 Tenacity k=5 VAT k=4 kNN3 Integrity ak=5 VAT k=2 Integrity k=5 Tenacity k=5
Silhouette 0.11 0.05 0.07 0.07 0.12 0.04 0.05
Davies-Bouldin 3.18 4.28 4.19 4.40 3.37 3.66 3.75
Xie-Beni 3.38 7.16 8.10 8.92 3.01 5.77 6.48
Dunn 0.13 0.15 0.15 0.17 0.14 0.14 0.14
Calinski-Harabasz 152.57 154.11 166.22 165.32 167.71 142.52 141.58
I Index 0.14 0.08 0.12 0.08 0.12 0.06 0.09
SD Index 9.96 14.62 14.40 20.10 8.52 7.71 9.10
SDb w Index 1.37 1.07 1.16 1.06 1.87 1.10 1.05
CVNN Index 1.38 0.95 1.21 0.54 2.00 2.00 2.00
Separability 31.63 11.14 20.21 8.34 8.53 11.39 15.25
Modularity (> 0.6) 0.42 0.72 0.65 0.68 0.27 0.67 0.68
Conductance (< 0.07) 0.02 0.04 0.03 0.06 0.06 0.05 0.04
  1. akNN3 using Integrity measure on correlation filtered data