Skip to main content

Table 3 Effectiveness of clustering of MDDR dataset using F-Measure: ECFP_4 Fingerprint

From: Voting-based consensus clustering for combining multiple clusterings of chemical structures

Clustering method No. of clusters
500 600 700 800 900 1000
Consensus clustering CVAA Correlation 33.58 29.81 24.44 20.09 18.41 17.43
Cosine 34.75 31.32 24.97 20.26 18.46 17.73
Euclidean 25.43 23.34 20.51 19.13 16.47 14.64
Hamming 25.48 24.04 20.23 19.62 17.31 14.73
Jaccard 35.71 33.17 28.66 21.8 19.63 18.86
Manhattan 25.41 23.98 20.30 19.53 17.25 14.65
CSPA Correlation 5.53 4.88 4.23 3.85 3.6 3.18
Cosine 5.43 4.88 4.28 3.91 3.55 3.10
Euclidean 5.47 4.87 4.17 3.79 3.53 3.33
Hamming 5.45 4.82 4.23 3.87 3.58 3.19
Jaccard 5.51 4.99 4.25 3.99 3.62 3.20
Manhattan 5.44 4.85 4.23 3.89 3.62 3.20
HGPA Correlation 7.01 6.2 5.21 4.5 4.16 3.68
Cosine 6.83 5.95 5.29 4.47 4.21 3.93
Euclidean 7.29 5.82 5.29 4.39 4.48 3.94
Hamming 7.01 5.83 5.29 4.50 4.37 3.69
Jaccard 6.87 5.91 5.31 4.81 4.80 3.66
Manhattan 7.81 5.17 5.38 4.61 4.66 3.68
Individual clustering Ward's method   11.61 10.71 9.04 8.29 7.64 7.02
\