|
||||||||||
| PREV PACKAGE NEXT PACKAGE | FRAMES NO FRAMES | |||||||||
See:
Description
| Class Summary | |
|---|---|
| CamberraNumericalDistance | The Camberra distance measure. |
| ChebychevNumericalDistance | This measure returns the maximal individual absolute distance of both examples in any component. |
| CorrelationSimilarity | Similarity based on the correlation coefficient. |
| CosineSimilarity | Cosine similarity that supports feature weights. |
| DiceNumericalSimilarity | A variant of the Dice coefficient defined for numeric attributes. |
| DTWDistance | A distance measure based on "Dynamic Time Warping". |
| EuclideanDistance | The euclidean distance. |
| InnerProductSimilarity | Similarity based on the inner product. |
| JaccardNumericalSimilarity | A variant of the Jaccard coefficient defined for numeric attributes. |
| KernelEuclideanDistance | This class uses the approach of Schoelkopf (2001) The Kernel Trick for Distances. |
| ManhattanDistance | The Manhattan distance. |
| MaxProductSimilarity | Specialized similarity that takes the maximum product of two feature values. |
| OverlapNumericalSimilarity | A variant of simple matching for numerical attributes. |
This package consists of similariy functions for numerical values only.
|
|
|||||||||
| PREV PACKAGE NEXT PACKAGE | FRAMES NO FRAMES | |||||||||