[MAHOUT-292] Classifier Test Data and Self Tests

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.3
• Fix Version/s: 0.3
• Assignee: Robin Anil
• Reporter: Robin Anil
• Created: 15/Feb/10 15:28
• Updated: 21/May/11 03:23
• Resolved: 15/Feb/10 18:28

Description

Till now there was no means to test if quality of classification suffered due to a code change.

Added Classifier data with 3 labels (mahout, lucene and spamassasin) with 4 long sentences in each of the labels.

Added a SelfTest which trains Bayes and CBayes model and classify the train dataset while testing and check accuracy and confusion matrix