Hidden Markov Model

Explain

You have three observable states {sleep, eat, poop} of your dog. 

From the past observations, you want to know the current state of your dog, {sick, healthy}

Since you don't know the current state, its hidden, therefore, hidden state.

To infer the hidden state, we need to know the following parameters

1) Initial probability for the hidden

2) Transition probability for the hidden

3) Emission probability, which is the probability of the observation given the hidden. For examples,

 

	Eat	Sleep	Poop
Sick	0.2	0.6	0.2
Healty	0.4	0.1	0.5

The problem is..How to estimate 1) ,2), and 3? 

1) and 2) can be obtained by the dynamic programming, 3) can be obtained by the EM-algorithm.

Visit, 

http://scikit-learn.sourceforge.net/stable/modules/hmm.html

http://www.blackarbs.com/blog/introduction-hidden-markov-models-python-networkx-sklearn/2/9/2017

Library installation (python 2.7)

1. install numpy related libraries 

python -m pip install --user numpy scipy matplotlib ipython jupyter pandas sympy nose

2. install scikit-learn

python -m pip install -U scikit-learn

3. install hmmlearn 

python -m pip install hmmlearn