A Decomposition of Classes via Clustering to Explain and Improve Naive Bayes.

Ricardo Vilalta Irina Rish

Published in: ECML (2003)

Keyphrases

naive bayes
decision trees
classification accuracy
logistic regression
text classification
naive bayes classifier
probability estimation
training data
text categorization

bayesian classifier
clustering algorithm
bayesian networks
uci datasets
unsupervised learning
independence assumption
classification algorithm
cost sensitive
augmented naive bayes

naive bayesian classifier
probabilistic classifiers
naive bayes classification
text classifiers
test instances
base classifiers
support vector
feature selection

machine learning
natural language processing
data points
bayesian network classifiers
combining classifiers
feature extraction
locally weighted
neural network