Improve multinomial Naive Bayes classifier

[ercanse]

Ideas:

• Feature selection: select a certain proportion of words with highest information gain
• Normalize feature vectors to the average vector length observed in the data
• Investigate locally weighted learning

Resources:

• A Comparative Study on Feature Selection in Text Categorization
• Multinomial Naive Bayes for Text Categorization Revisited
• scikit-learn TfidfVectorizer