NOTE. The number of unique clustering objects for autoselection by the R algorithm must be greater than 5.
The ClusterCountSelectionType enumeration is used to determine the method of cluster count selection.
It is used by the following property:
| Value | Brief description | |
|
0 | Fixed. User-defined. |
|
1 | VarDisAutoSelection. Autoselection by the Variance Distance algorithm. The algorithm is based on calculating of variation of values of attributes in the sampling an is applicable only to fuzzy clustering. |
|
2 | RAlgorithmAutoSelection. Autoselection by the R algorithm. The algorithm is based on analysis of stability of alternative classifications for available modifications of the source sample. NOTE. The number of unique clustering objects for autoselection by the R algorithm must be greater than 5. |
See also: