An n-way Baysian classifier built as an extension of Classifier4J.
A Bayes classifier implementation that determines the partition, or subspace from among a N set of subspaces of a sample space Y, with the greatest relative probability of containing a given attribute set of word terms. The classifier returns an identification of the partition for the one subspace that is the best probabilistic match for containing a given attribute set of word terms.
Implemented in an Eclipse view plugin as part of a larger system to perform sentence classification.
- Extract classifier to a stand-alone library.
- Update Classifier4J dependency.
- Documentation: see BayesPartitionClassifier and PersistantWordsDataSource.
EPL License. See the accompanying License.txt file.
- Classifier4J-0.51.jar