Concept - Naive Bayes Classifier

(Stub created by chunk 02; chunk 01 (s.21-23) defines it. See the main note Concept - Naive Bayes.)

Bayes rule plus the conditional independence assumption: $h_{NB}=\arg {\max }_{h_i}p(h_i){\prod }_{j}p(d_j\mid h_i)$ (chunk 01, s.23).

From chunk 02 (s.31-60)

Per-class word probabilities come from word-count histograms (Concept - Word Likelihood); variant for counts: Concept - Multinomial Naive Bayes; zero counts: Concept - Zero Probability Problem; unseen words: Concept - Out of Vocabulary (OOV).

Source chunks: 03 Text Classification - Chunk 01 (s.1-30), 03 Text Classification - Chunk 02 (s.31-60)

Related

Concept - Naive Bayes, Concept - Conditional Independence, Concept - Maximum A Posteriori, Concept - Word Likelihood. Up: 03 Text Classification - Index