Algorithm

This section discusses the basic theory and use of HMMs. Although most aspects of our linear HMMs have been described previously [Krogh et al., 1994a], this section reexamines several operational features as the foundation for the experimental evaluation that follows.

• The hidden Markov model
• Estimation of the model
  • Prior distribution and regularization
  • Problems with local maxima
• Multiple sequence alignment
• Searching
• Model surgery
• Modeling domains and motifs
• Additional features