Skip to main content
. 2021 Jul 9;11:14214. doi: 10.1038/s41598-021-93244-2

Table 2.

Performance of BCALoD, DBcsan, DPC and K-mean. The bolded values are the best clustering quality indexes.

Algorithm AR ARI NMI HI Nc/Ns
Case 1 dataset BCALoD 0.9834 0.9924 0.9481 0.9848 8/8
DBSCAN 0.9367 0.9709 0.8807 0.9418 6/8
DPC 0.9796 0.9907 0.9350 0.9813 6/8
K-mean 0.8631 0.9351 0.7918 0.8701 4/8
Case 1 dataset (Noise-containing) BCALoD 0.9220 0.9674 0.8533 0.9348 8/8
DBSCAN 0.8689 0.9444 0.7923 0.8888 7/8
DPC 0.8536 0.9374 0.7767 0.8748 6/8
K-mean 0.7639 0.8952 0.6616 0.7904 4/8
Case 2 dataset BCALoD 1 1 1 1 4/4
DBSCAN 1 1 1 1 4/4
DPC 1 1 1 1 4/4
K-mean 0.4293 0.7928 0.3792 0.5855 6/4