public class WEKAPipelineCharacterizer extends java.lang.Object implements IPipelineCharacterizer
| Constructor and Description |
|---|
WEKAPipelineCharacterizer() |
| Modifier and Type | Method and Description |
|---|---|
void |
build(java.util.List<MLPipeline> pipelines)
Finds frequent patterns in the given list of pipelines.
|
double[] |
characterize(MLPipeline pipeline)
Checks which of the found patterns (found during the training phase in
IPipelineCharacterizer.build(List)) occur in this pipeline. |
double[][] |
getCharacterizationsOfTrainingExamples()
For each
MLPipeline that was used in the training, return which found
pattern (found during the training phase in
IPipelineCharacterizer.build(List)) occurs in which pipeline. |
int |
getMinSupport()
Get the minimum support required for a pattern to be considered frequent for
the tree mining algorithm.
|
IOntologyConnector |
getOntologyConnector() |
protected java.lang.String |
makeStringTreeRepresentation(MLPipeline pipeline)
Converts the given MLPipeline to a String representation of its components
using the ontology
|
void |
setMinSupport(int minSupport)
Set the minimum support required for a pattern to be considered frequent for
the tree mining algorithm.
|
void |
setOntologyConnector(IOntologyConnector ontologyConnector) |
public void build(java.util.List<MLPipeline> pipelines)
IPipelineCharacterizerbuild in interface IPipelineCharacterizerpipelines - The pipelines to go though for patternspublic double[] characterize(MLPipeline pipeline)
IPipelineCharacterizerIPipelineCharacterizer.build(List)) occur in this pipeline.
If in the returned list l, l[j]=1, pattern p occurs in this pipeline.
Otherwise l[j] and pattern j doesn't occur in this pipeline.characterize in interface IPipelineCharacterizerpipeline - The pipeline for which pattern occurrence is checkedprotected java.lang.String makeStringTreeRepresentation(MLPipeline pipeline)
pipeline - public double[][] getCharacterizationsOfTrainingExamples()
IPipelineCharacterizerMLPipeline that was used in the training, return which found
pattern (found during the training phase in
IPipelineCharacterizer.build(List)) occurs in which pipeline.
If in the returned matrix m, m[i][j]=1, pattern p occurs in training pipeline
i. Otherwise m[i][j] and pattern j doesn't occur in training pipeline i.getCharacterizationsOfTrainingExamples in interface IPipelineCharacterizerpublic IOntologyConnector getOntologyConnector()
public void setOntologyConnector(IOntologyConnector ontologyConnector)
ontologyConnector - the ontologyConnector to setpublic int getMinSupport()
public void setMinSupport(int minSupport)
minSupport - The minimum support a tree pattern must have to be considered
frequent