| Package | Description |
|---|---|
| org.openimaj.experiment.evaluation.cluster | |
| org.openimaj.experiment.evaluation.cluster.analyser |
This package contains classes that can be used to evaluate clustering performance.
|
| Class and Description |
|---|
| ClusterAnalyser |
| Class and Description |
|---|
| AdjustedRandomIndexAnalysis
The adjsuted random index as described by: http://faculty.washington.edu/kayee/pca/supp.pdf
|
| ClusterAnalyser |
| ClusterStatsAnalysis
Counts some overall statistics of cluster evaluation such as min, max and average cluster length
|
| DecisionAnalysis
Counting the number of true positives, true negatives, flase postitives and false negatives
one can produce various cluster quality metrics including the fscore and randindex
|
| DecisionClusterAnalyser
Gather the true positives, true negatives, false positives, false negatives for a
DecisionAnalysis instance |
| FScoreAnalysis
Uses a decision analysis to produce the random index result
|
| FullMEAnalysis
Sina Samangooei (ss@ecs.soton.ac.uk)
|
| NMIAnalysis
Measures normalised mutual information.
|
| PurityAnalysis
A measure of how pure each cluster is.
|
| RandomBaselineClusterAnalysis
The result of a
TestRandomBaselineClusterAnalyser which wraps the
baseline result and result of an AnalysisResult |
| RandomBaselineSMEAnalysis
Results of applying a RandomBaselineSMEAnalysis to evaluate clustering.
|
| RandomBaselineWrappable
An
AnalysisResult which can offer some score and thus
be compared to a random baseline |
| RandomIndexAnalysis
Uses a decision analysis to produce the random index result
|
| SimpleMEAnalysis
Result of applying a
SimpleMEClusterAnalyser. |