Natural Language Processing · TEE · read-chapter

Markov chains and Hidden Markov Models

Unit 4, pages 47–73

  1. Markov chains: read a transition table and multiply a path
  2. Hidden states, observations, transitions and emissions
  3. Estimate transition and emission probabilities from tagged data
  4. HMM path probabilities: three complete numerical examples
  5. POS HMM worked corpus: Mary, Jane, can, will and spot