From: A new algorithm to train hidden Markov models for biological sequences with partial labels
Symbols | Explanations |
---|---|
\(\theta\) | Hidden Markov model: \(\theta = (\pi , A,B)\) |
N | States’ number in hidden Markov model |
K | Symbolic Number in hidden Markov model |
A | Transition matrix with dimension \(N \times N\) |
\(a_{ij}\) | Probability of state i transition to state j |
B | Emission matrix with dimension \(N \times K\) |
\(b_{j}(k)\) | Probability of state j emitted from symbol k |
\(\pi\) | Initial probability of states with dimension \(N \times 1\) |
\(O^s\) | The sth sequence with length \(T^s\) |
\(X^{s}\) | State sequence of \(O^s\) |
m | Total number of sequences |