|
Class Summary |
| AbstractCompositeUFH |
A composite unknown feature handler is one in which different unknown feature handlers
are used under different conditions, as determined by chooseUnknownFeatureHandler. |
| AbstractUnknownFeatureHandler |
Some basic abstract implementations of an UnknownFeatureHandler are provided. |
| AggregateFeatureValueSet |
Keeps track of the aggregate of each feature, when it is known. |
| BayesianConsistentRegionAggregate |
Title: |
| DefaultDifference |
Provides a default difference between two features, when the value of at least
one of them is not known. |
| DefaultValue |
Provides a default value for an unknown feature, with which a difference is
calculated. |
| DependenceCompositeUFH |
A DependenceCompositeUFH is a CompositeUFH in which one UFH is used for "independent"
features, and another is used for "dependent" ones. |
| FullAggregate |
Provides a "default" value for an unknown feature, obtained from the average
value of that feature for all cases in the case base. |
| NNAggregate |
Provides a "default" value for an unknown feature, obtained from the average
value of that feature for "nearby" cases (that know that feature). |
| OracleRegionAggregate |
Title: |
| RegionAggregate |
Provides a "default" value for an unknown feature, obtained from the aggregate
value of that feature for all cases in the current class. |