A Novel Local Density Hierarchical Clustering Algorithm Based on Reverse Nearest Neighbors
Table 2
Results of the algorithm on synthetic data sets.
Algorithms
k
C
F1
AMI
ARI
NR (%)
k
C
F1
AMI
ARI
NR (%)
Pathbased
cluto-t4.8k
RNN_LDH
6
3
1
1
1
0.00
34
8
0.91
0.87
0.91
1.70
RNN_DSC
6
3
0.99
0.94
0.96
0.30
31
7
0.88
0.85
0.87
0.10
IS_DSC
10
3
1
1
1
50.70
63
4
0.84
0.79
0.81
16.60
ISB_DSC
5
4
0.94
0.82
0.88
4.70
21
7
0.68
0.74
0.55
0.90
ADPC
35
2
0.66
0.4
0.4
0.00
400
7
0.69
0.7
0.59
0.00
Compound
cluto-t5.8k
RNN_LDH
7
6
1
0.99
1
6.80
33
6
0.87
0.83
0.85
5.00
RNN_DSC
8
6
0.89
0.86
0.87
0.00
32
7
0.83
0.8
0.8
0.90
IS_DSC
10
5
1
1
1
31.60
62
7
0.99
0.98
0.99
20.50
ISB_DSC
8
8
0.97
0.92
0.97
1.80
39
12
0.85
0.85
0.84
2.10
ADPC
9
7
0.8
0.8
0.62
0.00
560
6
0.82
0.79
0.78
0.00
Flame
cluto-t7.10k
RNN_LDH
7
2
1
1
1
0.80
40
9
0.92
0.92
0.93
2.60
RNN_DSC
8
2
1
0.96
0.98
0.00
28
10
0.88
0.88
0.9
0.30
IS_DSC
4
2
0.7
0
−0.01
21.70
30
6
0.87
0.91
0.82
16.60
ISB_DSC
4
2
0.69
0.01
−0.01
1.30
18
18
0.86
0.87
0.83
1.50
ADPC
27
2
1
1
1
0.00
395
10
0.49
0.56
0.33
0.00
Dim1024
cluto-t8.8k
RNN_LDH
63
16
1
1
1
0.00
27
8
0.96
0.93
0.96
1.30
RNN_DSC
59
16
1
1
1
1.40
22
8
0.94
0.91
0.95
0.10
IS_DSC
63
16
1
1
1
39.60
30
4
0.68
0.78
0.56
12.00
ISB_DSC
58
16
1
1
1
6.30
10
14
0.98
0.96
0.98
2.10
ADPC
2
16
1
1
1
0.00
240
9
0.59
0.6
0.45
0.00
Spiral
Jain
RNN_LDH
2
3
1
1
1
0.00
16
2
1
1
1
0.50
RNN_DSC
2
3
1
1
1
0.00
15
2
1
1
1
0.00
IS_DSC
5
3
1
1
1
50.30
16
2
1
1
1
36.20
ISB_DSC
2
3
1
1
1
0.00
16
2
1
1
1
0.00
ADPC
13
3
1
1
1
0.00
38
2
0.59
0.18
−0.02
0.00
If the number of clusters (C) found out is not correct, it is in italics. The best benchmark is written in bold in the condition of right cluster number.
