Interface defining probabilistic model with the viterbi, forward and backward algorithm. More...
#include <DecodableModel.hpp>
Inheritance diagram for tops::DecodableModel:
Public Member Functions | |
| virtual double | evaluate (const Sequence &s, unsigned int begin, unsigned int end) const |
| Calculates the sequence likelihood given this model. | |
| virtual double | forward (const Sequence &s, Matrix &alpha) const =0 |
| Forward algorithm. | |
| virtual double | backward (const Sequence &s, Matrix &beta) const =0 |
| Backward algorithm. | |
| virtual double | viterbi (const Sequence &s, Sequence &path, Matrix &gamma) const =0 |
| Viterbi algorithm. | |
| virtual void | choosePath (const Sequence &s, Sequence &path) |
| Choose a path given a sequence. | |
| virtual void | posteriorProbabilities (const Sequence &s, Matrix &probabilities) const |
| Posterior Probabilities: P(yi=k|x) | |
| virtual void | posteriorProbabilities (const Sequence &s, SparseMatrixPtr probabilities) const |
| Posterior Probabilities: P(yi=k|x) | |
| virtual void | posteriorDecoding (const Sequence &s, Sequence &path, Matrix &probabilities) const |
| Posterior Decoding: ^yi = argmax_k P(yi=k|x) | |
| virtual Sequence & | chooseObservation (Sequence &h, int i, int state) const =0 |
| Choose the observation given a state. | |
| virtual int | chooseState (int state) const =0 |
| Choose a state. | |
| virtual int | chooseFirstState () const =0 |
| Choose the initial state. | |
| virtual std::string | getStateName (int state) const =0 |
| Get state name. | |
| virtual AlphabetPtr | getStateNames () const =0 |
| Get the state names. | |
Detailed Description
Interface defining probabilistic model with the viterbi, forward and backward algorithm.
Definition at line 38 of file DecodableModel.hpp.
Member Function Documentation
| virtual Sequence& tops::DecodableModel::chooseObservation | ( | Sequence & | h, |
| int | i, | ||
| int | state | ||
| ) | const [pure virtual] |
Choose the observation given a state.
- Parameters:
-
h is the history
Implemented in tops::GeneralizedHiddenMarkovModel, and tops::HiddenMarkovModel.
The documentation for this class was generated from the following files:
