基于改进局部密度的可扩展层次聚类算法

陈斌, 谢文波, 付勋, 张恒基, 王欣

Density⁃based scalable hierarchical clustering

Bin Chen, Wenbo Xie, Xun Fu, Hengji Zhang, Xin Wang

表2 本文算法和对比算法在UCI数据集上的聚类结果

Table 2 The clustering results of our algorithm and other algorithms on UCI datasets

数据集 算法 RI ARI NMI 数据集 算法 RI ARI NMI mfeat⁃fourier DBSC(opt=0.65) 0.8810 0.3680 0.5345 mfeat⁃karhune DBSC(opt=0.2) 0.9305 0.6261 0.7384 DBSC(0.7) 0.8308 0.3204 0.5163 DBSC(0.7) 0.9090 0.5385 0.6990 HAC 0.1818 0.0040 0.0816 HAC 0.5016 0.1030 0.3423 Affinity 0.7980 0.2159 0.4523 Affinity 0.8965 0.5086 0.6778 DPC 0.5036 0.0796 0.1897 DPC 0.7228 0.1766 0.3798 GRINCH 0.1988 0.0002 0.0159 GRINCH 0.2091 <0.0001 0.0233 HK⁃means 0.8562 0.3169 0.4557 HK⁃means 0.8959 0.3964 0.5316 Munec 0.8168 0.2581 0.4821 Munec 0.8729 0.3294 0.6209 RSC 0.8367 0.3195 0.5046 RSC 0.8830 0.5172 0.7110 SCC 0.1028 <0.0001 0.0028 SCC 0.4869 0.0261 0.1333 streaKHC 0.8619 0.3035 0.4220 streaKHC 0.8737 0.3539 0.4710 mfeat⁃zernike DBSC(opt=0.35) 0.9084 0.5174 0.6270 segment DBSC(opt=0.3) 0.8910 0.5584 0.6396 DBSC(0.7) 0.8729 0.4322 0.6009 DBSC(0.7) 0.8862 0.5434 0.6246 HAC 0.5733 0.1368 0.4016 HAC 0.5983 0.2394 0.5757 Affinity 0.8036 0.2592 0.4650 Affinity 0.8752 0.5090 0.6481 DPC 0.4934 0.0296 0.0617 DPC 0.3970 0.0122 0.0320 GRINCH 0.3800 0.0601 0.1636 GRINCH 0.5897 0.2385 0.3908 HK⁃means 0.8565 0.3195 0.4323 HK⁃means 0.8412 0.4165 0.5391 Munec 0.8293 0.2354 0.5365 Munec 0.8428 0.2052 0.5154 RSC 0.8893 0.4667 0.5903 RSC 0.7311 0.3043 0.5329 SCC 0.6082 0.0701 0.1829 SCC 0.6148 0.2139 0.3669 streaKHC 0.8609 0.3158 0.4358 streaKHC 0.8592 0.4625 0.5699 optdigits DBSC(opt=0.15) 0.9722 0.8401 0.8694 letter DBSC(opt=0.4) 0.9261 0.1355 0.4027 DBSC(0.7) 0.9576 0.7603 0.8373 DBSC(0.7) 0.9140 0.1303 0.3941 HAC 0.8956 0.5517 0.7341 HAC 0.7841 0.0608 0.3342 Affinity 0.9640 0.8010 0.8433 Affinity 0.9047 0.1283 0.3830 DPC 0.6551 0.1146 0.2730 DPC 0.0663 0.0001 0.0066 GRINCH 0.2754 0.0160 0.1001 GRINCH 0.2311 0.0017 0.0514 HK⁃means 0.9119 0.5374 0.6347 HK⁃means 0.8323 0.0713 0.2440 Munec 0.7739 0.1791 0.5729 Munec 0.5131 0.0092 0.2367 RSC 0.8623 0.4832 0.7013 RSC 0.9145 0.1498 0.3927 SCC 0.7108 0.1906 0.3654 SCC 0.3948 0.0059 0.0511 streaKHC 0.8946 0.4515 0.5507 streaKHC 0.9125 0.0836 0.2396 avila DBSC(opt=0.15) 0.7249 0.0929 0.1842 Mean DBSC(opt) 0.8906 0.4483 0.5708 DBSC(0.7) 0.7134 0.0643 0.1820 DBSC(0.7) 0.8691 0.3985 0.5527 HAC 0.2820 <0.0001 0.0328 HAC 0.5452 0.1557 0.3575 Affinity 0.7113 0.0528 0.1617 Affinity 0.8505 0.3535 0.5188 DPC 0.5009 0.0151 0.0814 DPC 0.4770 0.0611 0.1463 GRINCH 0.2967 0.0069 0.0420 GRINCH 0.3115 0.0461 0.1124 HK⁃means 0.2598 <0.0001 0.0199 HK⁃means 0.7791 0.2932 0.4082 Munec 0.5761 0.0480 0.1615 Munec 0.7464 0.1806 0.4466 RSC 0.5271 0.0224 0.0714 RSC 0.8063 0.3233 0.5006 SCC 0.2484 <0.0001 0.0284 SCC 0.4524 0.0716 0.1615 streaKHC 0.7224 0.0245 0.0886 streaKHC 0.8550 0.2850 0.3968