class Cadmium::Classifier::Viterbi

• Cadmium::Classifier::Viterbi
• Reference
• Object

Overview

TODO : add Kneser-Ney smoothing This is a Hidden Markov Model classifier which uses the Viterbi algorithm. It is efficient in predicting a state given prior observations matched to states. In NLP, it is often used to attribut POS tags to words of a text. As such it is used by Cadmium::POSTagger.

Included Modules

• Apatite

Defined in:

cadmium/classifier/viterbi.cr

Constructors

• .new(ngrams_size : Int32 = 3)

Instance Method Summary

• #classify(sequence_of_observations : Array(String)) : Hash(String, String)
• #emission_matrix : Matrix(Float64)
• #epsilon : Float64
• #label_count : Hash(String, Int32)
• #load_model(filename : String = "model.zip")
• #lookup_table : Hash(String, Int32)
• #ngram_label_count : Hash(Array(String), Int32)
• #ngrams_size : Int32

  property initial_probabilities

• #observation_space : Set(String)
• #predicted_states : Array(String)
• #prior_ngram_label_count : Hash(Array(String), Int32)
• #save_model(filename : String = "model.zip")
• #sequence_of_ngrams : Array(Array(Tuple(String, String)))
• #sequence_of_observations : Array(String)
• #sequence_of_prior_ngrams : Array(Array(Tuple(String, String)))
• #state_space : Set(String)
• #token_count : Hash(String, Int32)

  observation_space, state_space, initial_probabilities, sequence_of_observations, transition_matrix, emission_matrix

• #token_label_count : Hash(Tuple(String, String), Int32)
• #train(training_data : Array(Tuple(String, String)))
• #training_data : Array(Tuple(String, String))
• #transition_matrix : Matrix(Float64)

Constructor Detail

Instance Method Detail