[HIVEMALL-126] Max Entropy Classifier for Text classification

• Type: New Feature
• Status: In Progress
• Priority: Major
• Resolution: Unresolved
• Fix Version/s: 0.7.0
• Assignee: Makoto Yui
• Reporter: Makoto Yui
• Created: 04/Jul/17 06:16
• Updated: 25/Nov/19 17:07

Description

Max Entropy Classifier (a.k.a. Multi-nominal/Multiclass Logistic Regression) [1,2] is useful for Text classification.

We have an incoming pull request of Max Entropy Classifier that uses Apache OpenNLP internally.

[1] http://blog.datumbox.com/machine-learning-tutorial-the-max-entropy-text-classifier/
[2] https://en.wikipedia.org/wiki/Multinomial_logistic_regression

Need to check the quality of OpenNLP MaxEnt and it might be better to create own Multinominal Logistic Regression (e.g., based on smile).
https://github.com/haifengl/smile/blob/master/core/src/main/java/smile/classification/LogisticRegression.java

Issue Links

• links to: GitHub Pull Request #93