Model Selection
Abstract
From the definition of the HMM, we know that an HMM consists of two important factors: the number of hidden states and the emission distributions of observations. In this chapter we will present some methods to choose the number of states for a HMM, to detect the emission distributions for the given data, and to choose the sample size of the data. We will then present a formula to compare two HMMs. Finally, after defining the HMM's factors, we will present some evaluation indicators to evaluate the chosen HMM.
Fields of Study & Subjects
Bayesian Methods and Mixture Models
Speech Recognition and Synthesis
Control Systems and Identification
Publication details
- Journal / source
- Chapman & Hall/CRC eBooks
- Pages
- 79–84
- Publication date
- 4 March 2026
- Language
- EN
- Document type
- Book chapter