Databases Reference
In-Depth Information
Chess & COG
10
3
Chess
Cog
10
2
10
1
10
0
10
−1
0.2
0.4
0.6
0.8
1
weighted minimum density
Fig. 8.
Chess and COG run time comparison
10
8
Mushroom
Chess
COG
10
6
10
4
10
2
10
0
0
0.2
0.4
0.6
0.8
1
minimum weighted density
Fig. 9.
Total number of clusters being found
when there are 10,000 clusters being found. Attributes are numbered such
that more dense larger attributes have larger ids. The search starts from the
attribute that has the largest id (45 in this case), and ends when it finishes
attribute 0. From the figure we can see that when using weighted density, more
clusters in the less dense subspaces are returned. Close examination reveals
that using minimum density threshold, seven attributes are not included in
any clusters. On the other hand, using weighted density, all attributes belong
