| Previous | Table of Contents | Next |
A ClusteringAttributeUsage is a subclass of AttributeUsage to support attribute usages that are specific to clustering algorithms.
Superclasses
AttributeUsage
Attributes
attributeComparisonFunction
When two records are compared then either the distance of similarity is of interest. In both cases, the measures can be computed
by a combination of ’inner’ function and an ’outer’ function. The inner function compares two single field values and the
outer function computes
an aggregation over all fields.type:multiplicity: |
|||||
| AttributeComparisonFunction (absDiff | gaussSim | delta | | |||||
| equal | table) | |||||
| exactly one |
similarityScale
If the attributeComparisonFunction admits a value, then similarityScale is that value. Only valid for numerical attributes.
Null otherwise. type: Double multiplicity: exactly one
References
comparisonMatrix
The ComparisonMatrix attribute specifies a matrix used by a clustering algorithm. There are several kinds of matrices used
within clustering models; for example, to describe covariances and similarities.
class: CategoryMatrix defined by: ClusteringAttrUsageRefComparisonMatrix multiplicity: zero or one inverse: CategoryMatrix::attributeUsage