MODIFIED BAUM WELCH ALGORITHM ON HIDDEN MARKOV MODELS (Case Study: DNA sequences of Hylobates, Pongo, Gorilla, Homo sapiens, Pan)
In Hidden Markov Model (HMM), the estimation of parameter HMM (transition matrix, emission matrix, and prior probability) become interesting problem. The conventional Baum-Welch algorithm is a solution to resolve this issue. In fact, there are many high dimentional sequence of observations, such as...
Saved in:
| Main Author: | |
|---|---|
| Format: | Theses |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/12305 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | In Hidden Markov Model (HMM), the estimation of parameter HMM (transition matrix, emission matrix, and prior probability) become interesting problem. The conventional Baum-Welch algorithm is a solution to resolve this issue. In fact, there are many high dimentional sequence of observations, such as DNA sequences, that make estimation process take much time. Baum-Welch with modification with class specific, Gaussian Mixture Model, and Likelihood Ratio is considered to solve this problem. The observation sequences is classified by class specific. The Emission matrix is estimated using Gaussian Mixture Model. The Gaussian Mixture Model compounds by the hidden state and observation state. |
|---|