.TH "mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >" 3 "Tue Sep 9 2014" "Version 1.0.10" "MLPACK" \" -*- nroff -*-
.ad l
.nh
.SH NAME
mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy > \- 
.PP
This class contains methods which can fit a \fBGMM\fP to observations using the EM algorithm\&.  

.SH SYNOPSIS
.br
.PP
.SS "Public Member Functions"

.in +1c
.ti -1c
.RI "\fBEMFit\fP (const size_t \fBmaxIterations\fP=300, const double \fBtolerance\fP=1e-10, InitialClusteringType clusterer=InitialClusteringType(), CovarianceConstraintPolicy constraint=CovarianceConstraintPolicy())"
.br
.RI "\fIConstruct the \fBEMFit\fP object, optionally passing an InitialClusteringType object (just in case it needs to store state)\&. \fP"
.ti -1c
.RI "const InitialClusteringType & \fBClusterer\fP () const "
.br
.RI "\fIGet the clusterer\&. \fP"
.ti -1c
.RI "InitialClusteringType & \fBClusterer\fP ()"
.br
.RI "\fIModify the clusterer\&. \fP"
.ti -1c
.RI "const CovarianceConstraintPolicy & \fBConstraint\fP () const "
.br
.RI "\fIGet the covariance constraint policy class\&. \fP"
.ti -1c
.RI "CovarianceConstraintPolicy & \fBConstraint\fP ()"
.br
.RI "\fIModify the covariance constraint policy class\&. \fP"
.ti -1c
.RI "void \fBEstimate\fP (const arma::mat &observations, std::vector< arma::vec > &means, std::vector< arma::mat > &covariances, arma::vec &weights, const bool useInitialModel=false)"
.br
.RI "\fIFit the observations to a Gaussian mixture model (\fBGMM\fP) using the EM algorithm\&. \fP"
.ti -1c
.RI "void \fBEstimate\fP (const arma::mat &observations, const arma::vec &probabilities, std::vector< arma::vec > &means, std::vector< arma::mat > &covariances, arma::vec &weights, const bool useInitialModel=false)"
.br
.RI "\fIFit the observations to a Gaussian mixture model (\fBGMM\fP) using the EM algorithm, taking into account the probabilities of each point being from this mixture\&. \fP"
.ti -1c
.RI "size_t \fBMaxIterations\fP () const "
.br
.RI "\fIGet the maximum number of iterations of the EM algorithm\&. \fP"
.ti -1c
.RI "size_t & \fBMaxIterations\fP ()"
.br
.RI "\fIModify the maximum number of iterations of the EM algorithm\&. \fP"
.ti -1c
.RI "double \fBTolerance\fP () const "
.br
.RI "\fIGet the tolerance for the convergence of the EM algorithm\&. \fP"
.ti -1c
.RI "double & \fBTolerance\fP ()"
.br
.RI "\fIModify the tolerance for the convergence of the EM algorithm\&. \fP"
.in -1c
.SS "Private Member Functions"

.in +1c
.ti -1c
.RI "void \fBInitialClustering\fP (const arma::mat &observations, std::vector< arma::vec > &means, std::vector< arma::mat > &covariances, arma::vec &weights)"
.br
.RI "\fIRun the clusterer, and then turn the cluster assignments into Gaussians\&. \fP"
.ti -1c
.RI "double \fBLogLikelihood\fP (const arma::mat &data, const std::vector< arma::vec > &means, const std::vector< arma::mat > &covariances, const arma::vec &weights) const "
.br
.RI "\fICalculate the log-likelihood of a model\&. \fP"
.in -1c
.SS "Private Attributes"

.in +1c
.ti -1c
.RI "InitialClusteringType \fBclusterer\fP"
.br
.RI "\fIObject which will perform the clustering\&. \fP"
.ti -1c
.RI "CovarianceConstraintPolicy \fBconstraint\fP"
.br
.RI "\fIObject which applies constraints to the covariance matrix\&. \fP"
.ti -1c
.RI "size_t \fBmaxIterations\fP"
.br
.RI "\fIMaximum iterations of EM algorithm\&. \fP"
.ti -1c
.RI "double \fBtolerance\fP"
.br
.RI "\fITolerance for convergence of EM\&. \fP"
.in -1c
.SH "Detailed Description"
.PP 

.SS "template<typename InitialClusteringType = kmeans::KMeans<>, typename CovarianceConstraintPolicy = PositiveDefiniteConstraint>class mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >"
This class contains methods which can fit a \fBGMM\fP to observations using the EM algorithm\&. 

It requires an initial clustering mechanism, which is by default the KMeans algorithm\&. The clustering mechanism must implement the following method:
.PP
.IP "\(bu" 2
void Cluster(const arma::mat& observations, const size_t clusters, arma::Col<size_t>& assignments);
.PP
.PP
This method should create 'clusters' clusters, and return the assignment of each point to a cluster\&. 
.PP
Definition at line 51 of file em_fit\&.hpp\&.
.SH "Constructor & Destructor Documentation"
.PP 
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::\fBEMFit\fP (const size_tmaxIterations = \fC300\fP, const doubletolerance = \fC1e-10\fP, InitialClusteringTypeclusterer = \fCInitialClusteringType()\fP, CovarianceConstraintPolicyconstraint = \fCCovarianceConstraintPolicy()\fP)"

.PP
Construct the \fBEMFit\fP object, optionally passing an InitialClusteringType object (just in case it needs to store state)\&. Setting the maximum number of iterations to 0 means that the EM algorithm will iterate until convergence (with the given tolerance)\&.
.PP
The parameter forcePositive controls whether or not the covariance matrices are checked for positive definiteness at each iteration\&. This could be a time-consuming task, so, if you know your data is well-behaved, you can set it to false and save some runtime\&.
.PP
\fBParameters:\fP
.RS 4
\fImaxIterations\fP Maximum number of iterations for EM\&. 
.br
\fItolerance\fP Log-likelihood tolerance required for convergence\&. 
.br
\fIforcePositive\fP Check for positive-definiteness of each covariance matrix at each iteration\&. 
.br
\fIclusterer\fP Object which will perform the initial clustering\&. 
.RE
.PP

.SH "Member Function Documentation"
.PP 
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> const InitialClusteringType& \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::Clusterer () const\fC [inline]\fP"

.PP
Get the clusterer\&. 
.PP
Definition at line 122 of file em_fit\&.hpp\&.
.PP
References mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::clusterer\&.
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> InitialClusteringType& \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::Clusterer ()\fC [inline]\fP"

.PP
Modify the clusterer\&. 
.PP
Definition at line 124 of file em_fit\&.hpp\&.
.PP
References mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::clusterer\&.
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> const CovarianceConstraintPolicy& \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::Constraint () const\fC [inline]\fP"

.PP
Get the covariance constraint policy class\&. 
.PP
Definition at line 127 of file em_fit\&.hpp\&.
.PP
References mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::constraint\&.
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> CovarianceConstraintPolicy& \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::Constraint ()\fC [inline]\fP"

.PP
Modify the covariance constraint policy class\&. 
.PP
Definition at line 129 of file em_fit\&.hpp\&.
.PP
References mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::constraint\&.
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> void \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::Estimate (const arma::mat &observations, std::vector< arma::vec > &means, std::vector< arma::mat > &covariances, arma::vec &weights, const booluseInitialModel = \fCfalse\fP)"

.PP
Fit the observations to a Gaussian mixture model (\fBGMM\fP) using the EM algorithm\&. The size of the vectors (indicating the number of components) must already be set\&. Optionally, if useInitialModel is set to true, then the model given in the means, covariances, and weights parameters is used as the initial model, instead of using the InitialClusteringType::Cluster() option\&.
.PP
\fBParameters:\fP
.RS 4
\fIobservations\fP List of observations to train on\&. 
.br
\fImeans\fP Vector to store trained means in\&. 
.br
\fIcovariances\fP Vector to store trained covariances in\&. 
.br
\fIweights\fP Vector to store a priori weights in\&. 
.br
\fIuseInitialModel\fP If true, the given model is used for the initial clustering\&. 
.RE
.PP

.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> void \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::Estimate (const arma::mat &observations, const arma::vec &probabilities, std::vector< arma::vec > &means, std::vector< arma::mat > &covariances, arma::vec &weights, const booluseInitialModel = \fCfalse\fP)"

.PP
Fit the observations to a Gaussian mixture model (\fBGMM\fP) using the EM algorithm, taking into account the probabilities of each point being from this mixture\&. The size of the vectors (indicating the number of components) must already be set\&. Optionally, if useInitialModel is set to true, then the model given in the means, covariances, and weights parameters is used as the initial model, instead of using the InitialClusteringType::Cluster() option\&.
.PP
\fBParameters:\fP
.RS 4
\fIobservations\fP List of observations to train on\&. 
.br
\fIprobabilities\fP Probability of each point being from this model\&. 
.br
\fImeans\fP Vector to store trained means in\&. 
.br
\fIcovariances\fP Vector to store trained covariances in\&. 
.br
\fIweights\fP Vector to store a priori weights in\&. 
.br
\fIuseInitialModel\fP If true, the given model is used for the initial clustering\&. 
.RE
.PP

.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> void \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::InitialClustering (const arma::mat &observations, std::vector< arma::vec > &means, std::vector< arma::mat > &covariances, arma::vec &weights)\fC [private]\fP"

.PP
Run the clusterer, and then turn the cluster assignments into Gaussians\&. This is a helper function for both overloads of \fBEstimate()\fP\&. The vectors must be already set to the number of clusters\&.
.PP
\fBParameters:\fP
.RS 4
\fIobservations\fP List of observations\&. 
.br
\fImeans\fP Vector to store means in\&. 
.br
\fIcovariances\fP Vector to store covariances in\&. 
.br
\fIweights\fP Vector to store a priori weights in\&. 
.RE
.PP

.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> double \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::LogLikelihood (const arma::mat &data, const std::vector< arma::vec > &means, const std::vector< arma::mat > &covariances, const arma::vec &weights) const\fC [private]\fP"

.PP
Calculate the log-likelihood of a model\&. Yes, this is reimplemented in the \fBGMM\fP code\&. Intuition suggests that the log-likelihood is not the best way to determine if the EM algorithm has converged\&.
.PP
\fBParameters:\fP
.RS 4
\fIdata\fP Data matrix\&. 
.br
\fImeans\fP Vector of means\&. 
.br
\fIcovariances\fP Vector of covariance matrices\&. 
.br
\fIweights\fP Vector of a priori weights\&. 
.RE
.PP

.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> size_t \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::MaxIterations () const\fC [inline]\fP"

.PP
Get the maximum number of iterations of the EM algorithm\&. 
.PP
Definition at line 132 of file em_fit\&.hpp\&.
.PP
References mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::maxIterations\&.
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> size_t& \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::MaxIterations ()\fC [inline]\fP"

.PP
Modify the maximum number of iterations of the EM algorithm\&. 
.PP
Definition at line 134 of file em_fit\&.hpp\&.
.PP
References mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::maxIterations\&.
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> double \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::Tolerance () const\fC [inline]\fP"

.PP
Get the tolerance for the convergence of the EM algorithm\&. 
.PP
Definition at line 137 of file em_fit\&.hpp\&.
.PP
References mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::tolerance\&.
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> double& \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::Tolerance ()\fC [inline]\fP"

.PP
Modify the tolerance for the convergence of the EM algorithm\&. 
.PP
Definition at line 139 of file em_fit\&.hpp\&.
.PP
References mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::tolerance\&.
.SH "Member Data Documentation"
.PP 
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> InitialClusteringType \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::clusterer\fC [private]\fP"

.PP
Object which will perform the clustering\&. 
.PP
Definition at line 177 of file em_fit\&.hpp\&.
.PP
Referenced by mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::Clusterer()\&.
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> CovarianceConstraintPolicy \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::constraint\fC [private]\fP"

.PP
Object which applies constraints to the covariance matrix\&. 
.PP
Definition at line 179 of file em_fit\&.hpp\&.
.PP
Referenced by mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::Constraint()\&.
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> size_t \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::maxIterations\fC [private]\fP"

.PP
Maximum iterations of EM algorithm\&. 
.PP
Definition at line 173 of file em_fit\&.hpp\&.
.PP
Referenced by mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::MaxIterations()\&.
.SS "template<typename InitialClusteringType  = kmeans::KMeans<>, typename CovarianceConstraintPolicy  = PositiveDefiniteConstraint> double \fBmlpack::gmm::EMFit\fP< InitialClusteringType, CovarianceConstraintPolicy >::tolerance\fC [private]\fP"

.PP
Tolerance for convergence of EM\&. 
.PP
Definition at line 175 of file em_fit\&.hpp\&.
.PP
Referenced by mlpack::gmm::EMFit< InitialClusteringType, CovarianceConstraintPolicy >::Tolerance()\&.

.SH "Author"
.PP 
Generated automatically by Doxygen for MLPACK from the source code\&.