[ https://issues.apache.org/jira/browse/SPARK-15906?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
MIN-FU YANG updated SPARK-15906: -------------------------------- Description: Improve the Naive Bayes algorithm on skew data according to "Tackling the Poor Assumptions of Naive Bayes Text Classifers" chapter 3.2 http://people.csail.mit.edu/jrennie/papers/icml03-nb.pdf Mahout & WEKA both have Complementary Naive Bayes implementations. https://mahout.apache.org/users/classification/bayesian.html http://weka.sourceforge.net/doc.packages/complementNaiveBayes/weka/classifiers/bayes/ComplementNaiveBayes.html Besides, this paper is referenced by other papers & books 600+ times, I think it's result is solid. https://scholar.google.com.tw/scholar?rlz=1C5CHFA_enTW567TW567&safe=high&um=1&ie=UTF-8&lr&cites=1197073324019480518 was: Improve the Naive Bayes algorithm on skew data according to "Tackling the Poor Assumptions of Naive Bayes Text Classifers" chapter 3.2 http://people.csail.mit.edu/jrennie/papers/icml03-nb.pdf Mahout & WEKA both have Complementary Naive Bayes implementations. https://mahout.apache.org/users/classification/bayesian.html http://weka.sourceforge.net/doc.packages/complementNaiveBayes/weka/classifiers/bayes/ComplementNaiveBayes.html Besides, this paper is referenced by other papers & books 600+ times, I think it's result is solid. > Complementary Naive Bayes Algorithm Implementation > -------------------------------------------------- > > Key: SPARK-15906 > URL: https://issues.apache.org/jira/browse/SPARK-15906 > Project: Spark > Issue Type: Improvement > Components: MLlib > Reporter: MIN-FU YANG > Priority: Minor > > Improve the Naive Bayes algorithm on skew data according to > "Tackling the Poor Assumptions of Naive Bayes Text Classifers" chapter 3.2 > http://people.csail.mit.edu/jrennie/papers/icml03-nb.pdf > Mahout & WEKA both have Complementary Naive Bayes implementations. > https://mahout.apache.org/users/classification/bayesian.html > http://weka.sourceforge.net/doc.packages/complementNaiveBayes/weka/classifiers/bayes/ComplementNaiveBayes.html > Besides, this paper is referenced by other papers & books 600+ times, I think > it's result is solid. > https://scholar.google.com.tw/scholar?rlz=1C5CHFA_enTW567TW567&safe=high&um=1&ie=UTF-8&lr&cites=1197073324019480518 -- This message was sent by Atlassian JIRA (v6.3.4#6332) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org