Mahee 718&saba 710
Mahee 718&saba 710
Mahee 718&saba 710
Submitted by:
Maheen Fatima(718)
Saba Tasneem (710)
Submitted to:
Department of Bioinformatics
Faculty of Computing
International Islamic University Islamabad
Hidden Markov Model
Definition:
A Hidden Markov Model (HMM) is a statistical model used to represent systems with
unobserved (hidden) states that generate a sequence of observable events
(observations).
It assumes that the system transitions between hidden states probabilistically, and
each state has a distinct probability of producing each possible observation.
These models work especially well with sequential data, where the underlying states
output observable data but are not directly observable.
Components:
States:
Hidden States:
These represent the unobservable, underlying states of the system. For example,
hidden states might represent weather conditions in a weather forecasting application.
Observable States:
Observable states are the states that emit data. In the weather example, observable
states could represent observed weather parameters like temperature and humidity.
Transition Probabilities:
Define the likelihood of transitioning from one hidden state to another. These
probabilities capture the dynamics of the system.
Emission Probabilities:
Determine the likelihood of an observable state emitting specific data. In our weather
example, emission probabilities would describe the chance of observing particular
weather conditions given the hidden state.
Application and Usage:
Speech Recognition: To accurately convert speech to text, speech recognition
systems make considerable use of HMMs to represent the transition between
phonemes.
Economics and Finance: To capture changes between various market states and aid
in the prediction of market trends, HMMs are used in financial market modeling.
Usage in Bioinformatics:
Gene Prediction:
HMMs are used for gene prediction, where hidden states represent coding and non-
coding regions, aiding in identifying potential genes in DNA sequences.
Phylogenetic Analysis:
HMMs are applied to model evolutionary processes, helping in constructing
phylogenetic trees and understanding the relationships between species.
Reference:
1. A Systematic Review of Hidden Markov Models and Their
Applications | Archives of Computational Methods in Engineering
(springer.com)
2. https://www.researchgate.net/figure/Graphical-representation-of-HMM-and-CHMM-
rolled-out-in-time-of-the-technical-details_fig4_2369443
3. Disaggregated Homes: The pros and cons of using HMMs to model appliances
(oliverparson.co.uk)
4. Advantages and disadvantages of hidden markov model | PPT
(slideshare.net)