C# Class Accord.Statistics.Models.Markov.Learning.BaseBaumWelchLearning

Base class for implementations of the Baum-Welch learning algorithm. This class cannot be instantiated.
Inheritance: BaseHiddenMarkovModelLearning, IUnsupervisedLearning, IConvergenceLearning
Show file Open project: accord-net/framework

Protected Methods

Method Description
BaseBaumWelchLearning ( IHiddenMarkovModel model ) : System

Initializes a new instance of the BaseBaumWelchLearning class.
ComputeForwardBackward ( int index, double lnFwd, double lnBwd ) : void

Computes the forward and backward probabilities matrices for a given observation referenced by its index in the input training data.
ComputeKsi ( int index, double lnFwd, double lnBwd ) : void

Computes the ksi matrix of probabilities for a given observation referenced by its index in the input training data.
Run ( Array observations ) : double

Runs the Baum-Welch learning algorithm for hidden Markov models.

Learning problem. Given some training observation sequences O = {o1, o2, ..., oK} and general structure of HMM (numbers of hidden and visible states), determine HMM parameters M = (A, B, pi) that best fit training data.

Run ( Array observations, double weights ) : double

Runs the Baum-Welch learning algorithm for hidden Markov models.

Learning problem. Given some training observation sequences O = {o1, o2, ..., oK} and general structure of HMM (numbers of hidden and visible states), determine HMM parameters M = (A, B, pi) that best fit training data.

Private Methods

Method Description
run ( Array observations ) : double

Method Details

BaseBaumWelchLearning() protected method

Initializes a new instance of the BaseBaumWelchLearning class.
protected BaseBaumWelchLearning ( IHiddenMarkovModel model ) : System
model IHiddenMarkovModel
return System

ComputeForwardBackward() protected abstract method

Computes the forward and backward probabilities matrices for a given observation referenced by its index in the input training data.
protected abstract ComputeForwardBackward ( int index, double lnFwd, double lnBwd ) : void
index int The index of the observation in the input training data.
lnFwd double Returns the computed forward probabilities matrix.
lnBwd double Returns the computed backward probabilities matrix.
return void

ComputeKsi() protected abstract method

Computes the ksi matrix of probabilities for a given observation referenced by its index in the input training data.
protected abstract ComputeKsi ( int index, double lnFwd, double lnBwd ) : void
index int The index of the observation in the input training data.
lnFwd double The matrix of forward probabilities for the observation.
lnBwd double The matrix of backward probabilities for the observation.
return void

Run() protected method

Runs the Baum-Welch learning algorithm for hidden Markov models.
Learning problem. Given some training observation sequences O = {o1, o2, ..., oK} and general structure of HMM (numbers of hidden and visible states), determine HMM parameters M = (A, B, pi) that best fit training data.
protected Run ( Array observations ) : double
observations System.Array /// The sequences of univariate or multivariate observations used to train the model. /// Can be either of type double[] (for the univariate case) or double[][] for the /// multivariate case. ///
return double

Run() protected method

Runs the Baum-Welch learning algorithm for hidden Markov models.
Learning problem. Given some training observation sequences O = {o1, o2, ..., oK} and general structure of HMM (numbers of hidden and visible states), determine HMM parameters M = (A, B, pi) that best fit training data.
protected Run ( Array observations, double weights ) : double
observations System.Array /// The sequences of univariate or multivariate observations used to train the model. /// Can be either of type double[] (for the univariate case) or double[][] for the /// multivariate case.
weights double /// The weight associated with each sequence.
return double