classical-mlsupervisedprobabilisticnlp
Naive Bayes classifier
A generative probabilistic classifier assuming feature independence — surprisingly effective for text.
Уровни глубины
L0Intro~2ч
Applies Bayes' rule to a coin/spam example with 2 features.
L1Basics~8ч
Implements Gaussian/Multinomial/Bernoulli NB; handles Laplace smoothing.
L2Working~10ч
Builds a text classifier with TF-IDF + Multinomial NB; compares with logistic regression.
L3Advanced~20ч
Semi-supervised NB (EM); Bayesian networks as a generalisation.
L4Research~40ч
Tree-augmented NB, probabilistic graphical models research.