View Source Evision.ML.EM (Evision v0.2.9)

Summary

Types

t()

Type that represents an ML.EM struct.

Functions

Computes error on the training or test dataset

Computes error on the training or test dataset

Clears the algorithm state

create

getClustersNumber

getCovarianceMatrixType

Returns covariation matrices

Returns covariation matrices

Returns the cluster centers (means of the Gaussian mixture)

getTermCriteria

Returns the number of variables in training samples

Returns weights of the mixtures

Returns true if the model is classifier

Returns true if the model is trained

Loads and creates a serialized EM from a file

Loads and creates a serialized EM from a file

Returns a likelihood logarithm value and an index of the most probable mixture component for the given sample.

Returns a likelihood logarithm value and an index of the most probable mixture component for the given sample.

Returns posterior probabilities for the provided samples

Returns posterior probabilities for the provided samples

Reads algorithm parameters from a file storage

setClustersNumber

setCovarianceMatrixType

setTermCriteria

Trains the statistical model

Trains the statistical model

Trains the statistical model

Estimate the Gaussian mixture parameters from a samples set.

Estimate the Gaussian mixture parameters from a samples set.

Estimate the Gaussian mixture parameters from a samples set.

Estimate the Gaussian mixture parameters from a samples set.

Estimate the Gaussian mixture parameters from a samples set.

Estimate the Gaussian mixture parameters from a samples set.

Stores algorithm parameters in a file storage

Enumerator

Types

@type enum() :: integer()
@type t() :: %Evision.ML.EM{ref: reference()}

Type that represents an ML.EM struct.

  • ref. reference()

    The underlying erlang resource variable.

Functions

@spec calcError(Keyword.t()) :: any() | {:error, String.t()}
Link to this function

calcError(self, data, test)

View Source
@spec calcError(t(), Evision.ML.TrainData.t(), boolean()) ::
  {number(), Evision.Mat.t()} | {:error, String.t()}

Computes error on the training or test dataset

Positional Arguments
  • self: Evision.ML.EM.t()

  • data: Evision.ML.TrainData.t().

    the training data

  • test: bool.

    if true, the error is computed over the test subset of the data, otherwise it's computed over the training subset of the data. Please note that if you loaded a completely different dataset to evaluate already trained classifier, you will probably want not to set the test subset at all with TrainData::setTrainTestSplitRatio and specify test=false, so that the error is computed for the whole new set. Yes, this sounds a bit confusing.

Return
  • retval: float

  • resp: Evision.Mat.t().

    the optional output responses.

The method uses StatModel::predict to compute the error. For regression models the error is computed as RMS, for classifiers - as a percent of missclassified samples (0%-100%).

Python prototype (for reference only):

calcError(data, test[, resp]) -> retval, resp
Link to this function

calcError(self, data, test, opts)

View Source
@spec calcError(
  t(),
  Evision.ML.TrainData.t(),
  boolean(),
  [{atom(), term()}, ...] | nil
) ::
  {number(), Evision.Mat.t()} | {:error, String.t()}

Computes error on the training or test dataset

Positional Arguments
  • self: Evision.ML.EM.t()

  • data: Evision.ML.TrainData.t().

    the training data

  • test: bool.

    if true, the error is computed over the test subset of the data, otherwise it's computed over the training subset of the data. Please note that if you loaded a completely different dataset to evaluate already trained classifier, you will probably want not to set the test subset at all with TrainData::setTrainTestSplitRatio and specify test=false, so that the error is computed for the whole new set. Yes, this sounds a bit confusing.

Return
  • retval: float

  • resp: Evision.Mat.t().

    the optional output responses.

The method uses StatModel::predict to compute the error. For regression models the error is computed as RMS, for classifiers - as a percent of missclassified samples (0%-100%).

Python prototype (for reference only):

calcError(data, test[, resp]) -> retval, resp
@spec clear(Keyword.t()) :: any() | {:error, String.t()}
@spec clear(t()) :: t() | {:error, String.t()}

Clears the algorithm state

Positional Arguments
  • self: Evision.ML.EM.t()

Python prototype (for reference only):

clear() -> None
@spec create() :: t() | {:error, String.t()}

create

Return
  • retval: Evision.ML.EM.t()

Creates empty %EM model. The model should be trained then using StatModel::train(traindata, flags) method. Alternatively, you can use one of the EM::train* methods or load it from file using Algorithm::load\<EM>(filename).

Python prototype (for reference only):

create() -> retval
@spec create(Keyword.t()) :: any() | {:error, String.t()}
@spec empty(Keyword.t()) :: any() | {:error, String.t()}
@spec empty(t()) :: boolean() | {:error, String.t()}

empty

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • retval: bool

Python prototype (for reference only):

empty() -> retval
Link to this function

getClustersNumber(named_args)

View Source
@spec getClustersNumber(Keyword.t()) :: any() | {:error, String.t()}
@spec getClustersNumber(t()) :: integer() | {:error, String.t()}

getClustersNumber

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • retval: integer()

@see setClustersNumber/2

Python prototype (for reference only):

getClustersNumber() -> retval
Link to this function

getCovarianceMatrixType(named_args)

View Source
@spec getCovarianceMatrixType(Keyword.t()) :: any() | {:error, String.t()}
@spec getCovarianceMatrixType(t()) :: integer() | {:error, String.t()}

getCovarianceMatrixType

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • retval: integer()

@see setCovarianceMatrixType/2

Python prototype (for reference only):

getCovarianceMatrixType() -> retval
@spec getCovs(Keyword.t()) :: any() | {:error, String.t()}
@spec getCovs(t()) :: [Evision.Mat.t()] | {:error, String.t()}

Returns covariation matrices

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • covs: [Evision.Mat].

Returns vector of covariation matrices. Number of matrices is the number of gaussian mixtures, each matrix is a square floating-point matrix NxN, where N is the space dimensionality.

Python prototype (for reference only):

getCovs([, covs]) -> covs
@spec getCovs(t(), [{atom(), term()}, ...] | nil) ::
  [Evision.Mat.t()] | {:error, String.t()}

Returns covariation matrices

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • covs: [Evision.Mat].

Returns vector of covariation matrices. Number of matrices is the number of gaussian mixtures, each matrix is a square floating-point matrix NxN, where N is the space dimensionality.

Python prototype (for reference only):

getCovs([, covs]) -> covs
Link to this function

getDefaultName(named_args)

View Source
@spec getDefaultName(Keyword.t()) :: any() | {:error, String.t()}
@spec getDefaultName(t()) :: binary() | {:error, String.t()}

getDefaultName

Positional Arguments
  • self: Evision.ML.EM.t()
Return

Returns the algorithm string identifier. This string is used as top level xml/yml node tag when the object is saved to a file or string.

Python prototype (for reference only):

getDefaultName() -> retval
@spec getMeans(Keyword.t()) :: any() | {:error, String.t()}
@spec getMeans(t()) :: Evision.Mat.t() | {:error, String.t()}

Returns the cluster centers (means of the Gaussian mixture)

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • retval: Evision.Mat.t()

Returns matrix with the number of rows equal to the number of mixtures and number of columns equal to the space dimensionality.

Python prototype (for reference only):

getMeans() -> retval
Link to this function

getTermCriteria(named_args)

View Source
@spec getTermCriteria(Keyword.t()) :: any() | {:error, String.t()}
@spec getTermCriteria(t()) :: {integer(), integer(), number()} | {:error, String.t()}

getTermCriteria

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • retval: TermCriteria

@see setTermCriteria/2

Python prototype (for reference only):

getTermCriteria() -> retval
@spec getVarCount(Keyword.t()) :: any() | {:error, String.t()}
@spec getVarCount(t()) :: integer() | {:error, String.t()}

Returns the number of variables in training samples

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • retval: integer()

Python prototype (for reference only):

getVarCount() -> retval
@spec getWeights(Keyword.t()) :: any() | {:error, String.t()}
@spec getWeights(t()) :: Evision.Mat.t() | {:error, String.t()}

Returns weights of the mixtures

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • retval: Evision.Mat.t()

Returns vector with the number of elements equal to the number of mixtures.

Python prototype (for reference only):

getWeights() -> retval
Link to this function

isClassifier(named_args)

View Source
@spec isClassifier(Keyword.t()) :: any() | {:error, String.t()}
@spec isClassifier(t()) :: boolean() | {:error, String.t()}

Returns true if the model is classifier

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • retval: bool

Python prototype (for reference only):

isClassifier() -> retval
@spec isTrained(Keyword.t()) :: any() | {:error, String.t()}
@spec isTrained(t()) :: boolean() | {:error, String.t()}

Returns true if the model is trained

Positional Arguments
  • self: Evision.ML.EM.t()
Return
  • retval: bool

Python prototype (for reference only):

isTrained() -> retval
@spec load(Keyword.t()) :: any() | {:error, String.t()}
@spec load(binary()) :: t() | {:error, String.t()}

Loads and creates a serialized EM from a file

Positional Arguments
  • filepath: String.

    path to serialized EM

Keyword Arguments
  • nodeName: String.

    name of node containing the classifier

Return
  • retval: Evision.ML.EM.t()

Use EM::save to serialize and store an EM to disk. Load the EM from this file again, by calling this function with the path to the file. Optionally specify the node for the file containing the classifier

Python prototype (for reference only):

load(filepath[, nodeName]) -> retval
@spec load(binary(), [{:nodeName, term()}] | nil) :: t() | {:error, String.t()}

Loads and creates a serialized EM from a file

Positional Arguments
  • filepath: String.

    path to serialized EM

Keyword Arguments
  • nodeName: String.

    name of node containing the classifier

Return
  • retval: Evision.ML.EM.t()

Use EM::save to serialize and store an EM to disk. Load the EM from this file again, by calling this function with the path to the file. Optionally specify the node for the file containing the classifier

Python prototype (for reference only):

load(filepath[, nodeName]) -> retval
@spec predict2(Keyword.t()) :: any() | {:error, String.t()}
@spec predict2(t(), Evision.Mat.maybe_mat_in()) ::
  {{number(), number()}, Evision.Mat.t()} | {:error, String.t()}

Returns a likelihood logarithm value and an index of the most probable mixture component for the given sample.

Positional Arguments
  • self: Evision.ML.EM.t()

  • sample: Evision.Mat.

    A sample for classification. It should be a one-channel matrix of \f$1 \times dims\f$ or \f$dims \times 1\f$ size.

Return
  • retval: Vec2d

  • probs: Evision.Mat.t().

    Optional output matrix that contains posterior probabilities of each component given the sample. It has \f$1 \times nclusters\f$ size and CV_64FC1 type.

The method returns a two-element double vector. Zero element is a likelihood logarithm value for the sample. First element is an index of the most probable mixture component for the given sample.

Python prototype (for reference only):

predict2(sample[, probs]) -> retval, probs
Link to this function

predict2(self, sample, opts)

View Source
@spec predict2(t(), Evision.Mat.maybe_mat_in(), [{atom(), term()}, ...] | nil) ::
  {{number(), number()}, Evision.Mat.t()} | {:error, String.t()}

Returns a likelihood logarithm value and an index of the most probable mixture component for the given sample.

Positional Arguments
  • self: Evision.ML.EM.t()

  • sample: Evision.Mat.

    A sample for classification. It should be a one-channel matrix of \f$1 \times dims\f$ or \f$dims \times 1\f$ size.

Return
  • retval: Vec2d

  • probs: Evision.Mat.t().

    Optional output matrix that contains posterior probabilities of each component given the sample. It has \f$1 \times nclusters\f$ size and CV_64FC1 type.

The method returns a two-element double vector. Zero element is a likelihood logarithm value for the sample. First element is an index of the most probable mixture component for the given sample.

Python prototype (for reference only):

predict2(sample[, probs]) -> retval, probs
@spec predict(Keyword.t()) :: any() | {:error, String.t()}
@spec predict(t(), Evision.Mat.maybe_mat_in()) ::
  {number(), Evision.Mat.t()} | {:error, String.t()}

Returns posterior probabilities for the provided samples

Positional Arguments
  • self: Evision.ML.EM.t()

  • samples: Evision.Mat.

    The input samples, floating-point matrix

Keyword Arguments
  • flags: integer().

    This parameter will be ignored

Return
  • retval: float

  • results: Evision.Mat.t().

    The optional output \f$ nSamples \times nClusters\f$ matrix of results. It contains posterior probabilities for each sample from the input

Python prototype (for reference only):

predict(samples[, results[, flags]]) -> retval, results
Link to this function

predict(self, samples, opts)

View Source
@spec predict(t(), Evision.Mat.maybe_mat_in(), [{:flags, term()}] | nil) ::
  {number(), Evision.Mat.t()} | {:error, String.t()}

Returns posterior probabilities for the provided samples

Positional Arguments
  • self: Evision.ML.EM.t()

  • samples: Evision.Mat.

    The input samples, floating-point matrix

Keyword Arguments
  • flags: integer().

    This parameter will be ignored

Return
  • retval: float

  • results: Evision.Mat.t().

    The optional output \f$ nSamples \times nClusters\f$ matrix of results. It contains posterior probabilities for each sample from the input

Python prototype (for reference only):

predict(samples[, results[, flags]]) -> retval, results
@spec read(Keyword.t()) :: any() | {:error, String.t()}
@spec read(t(), Evision.FileNode.t()) :: t() | {:error, String.t()}

Reads algorithm parameters from a file storage

Positional Arguments

Python prototype (for reference only):

read(fn) -> None
@spec save(Keyword.t()) :: any() | {:error, String.t()}
@spec save(t(), binary()) :: t() | {:error, String.t()}

save

Positional Arguments
  • self: Evision.ML.EM.t()
  • filename: String

Saves the algorithm to a file. In order to make this method work, the derived class must implement Algorithm::write(FileStorage& fs).

Python prototype (for reference only):

save(filename) -> None
Link to this function

setClustersNumber(named_args)

View Source
@spec setClustersNumber(Keyword.t()) :: any() | {:error, String.t()}
Link to this function

setClustersNumber(self, val)

View Source
@spec setClustersNumber(t(), integer()) :: t() | {:error, String.t()}

setClustersNumber

Positional Arguments
  • self: Evision.ML.EM.t()
  • val: integer()

@see getClustersNumber/1

Python prototype (for reference only):

setClustersNumber(val) -> None
Link to this function

setCovarianceMatrixType(named_args)

View Source
@spec setCovarianceMatrixType(Keyword.t()) :: any() | {:error, String.t()}
Link to this function

setCovarianceMatrixType(self, val)

View Source
@spec setCovarianceMatrixType(t(), integer()) :: t() | {:error, String.t()}

setCovarianceMatrixType

Positional Arguments
  • self: Evision.ML.EM.t()
  • val: integer()

@see getCovarianceMatrixType/1

Python prototype (for reference only):

setCovarianceMatrixType(val) -> None
Link to this function

setTermCriteria(named_args)

View Source
@spec setTermCriteria(Keyword.t()) :: any() | {:error, String.t()}
Link to this function

setTermCriteria(self, val)

View Source
@spec setTermCriteria(t(), {integer(), integer(), number()}) ::
  t() | {:error, String.t()}

setTermCriteria

Positional Arguments
  • self: Evision.ML.EM.t()
  • val: TermCriteria

@see getTermCriteria/1

Python prototype (for reference only):

setTermCriteria(val) -> None
@spec train(Keyword.t()) :: any() | {:error, String.t()}
@spec train(t(), Evision.ML.TrainData.t()) :: boolean() | {:error, String.t()}

Trains the statistical model

Positional Arguments
  • self: Evision.ML.EM.t()

  • trainData: Evision.ML.TrainData.t().

    training data that can be loaded from file using TrainData::loadFromCSV or created with TrainData::create.

Keyword Arguments
  • flags: integer().

    optional flags, depending on the model. Some of the models can be updated with the new training samples, not completely overwritten (such as NormalBayesClassifier or ANN_MLP).

Return
  • retval: bool

Python prototype (for reference only):

train(trainData[, flags]) -> retval
Link to this function

train(self, trainData, opts)

View Source
@spec train(t(), Evision.ML.TrainData.t(), [{:flags, term()}] | nil) ::
  boolean() | {:error, String.t()}

Trains the statistical model

Positional Arguments
  • self: Evision.ML.EM.t()

  • trainData: Evision.ML.TrainData.t().

    training data that can be loaded from file using TrainData::loadFromCSV or created with TrainData::create.

Keyword Arguments
  • flags: integer().

    optional flags, depending on the model. Some of the models can be updated with the new training samples, not completely overwritten (such as NormalBayesClassifier or ANN_MLP).

Return
  • retval: bool

Python prototype (for reference only):

train(trainData[, flags]) -> retval
Link to this function

train(self, samples, layout, responses)

View Source
@spec train(t(), Evision.Mat.maybe_mat_in(), integer(), Evision.Mat.maybe_mat_in()) ::
  boolean() | {:error, String.t()}

Trains the statistical model

Positional Arguments
  • self: Evision.ML.EM.t()

  • samples: Evision.Mat.

    training samples

  • layout: integer().

    See ml::SampleTypes.

  • responses: Evision.Mat.

    vector of responses associated with the training samples.

Return
  • retval: bool

Python prototype (for reference only):

train(samples, layout, responses) -> retval
@spec trainE(Keyword.t()) :: any() | {:error, String.t()}
Link to this function

trainE(self, samples, means0)

View Source
@spec trainE(t(), Evision.Mat.maybe_mat_in(), Evision.Mat.maybe_mat_in()) ::
  {Evision.Mat.t(), Evision.Mat.t(), Evision.Mat.t()}
  | false
  | {:error, String.t()}

Estimate the Gaussian mixture parameters from a samples set.

Positional Arguments
  • self: Evision.ML.EM.t()

  • samples: Evision.Mat.

    Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.

  • means0: Evision.Mat.

    Initial means \f$a_k\f$ of mixture components. It is a one-channel matrix of \f$nclusters \times dims\f$ size. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.

Keyword Arguments
  • covs0: Evision.Mat.

    The vector of initial covariance matrices \f$S_k\f$ of mixture components. Each of covariance matrices is a one-channel matrix of \f$dims \times dims\f$ size. If the matrices do not have CV_64F type they will be converted to the inner matrices of such type for the further computing.

  • weights0: Evision.Mat.

    Initial weights \f$\pi_k\f$ of mixture components. It should be a one-channel floating-point matrix with \f$1 \times nclusters\f$ or \f$nclusters \times 1\f$ size.

Return
  • retval: bool

  • logLikelihoods: Evision.Mat.t().

    The optional output matrix that contains a likelihood logarithm value for each sample. It has \f$nsamples \times 1\f$ size and CV_64FC1 type.

  • labels: Evision.Mat.t().

    The optional output "class label" for each sample: \f$\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\f$ (indices of the most probable mixture component for each sample). It has \f$nsamples \times 1\f$ size and CV_32SC1 type.

  • probs: Evision.Mat.t().

    The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \f$nsamples \times nclusters\f$ size and CV_64FC1 type.

This variation starts with Expectation step. You need to provide initial means \f$a\_k\f$ of mixture components. Optionally you can pass initial weights \f$\pi\_k\f$ and covariance matrices \f$S\_k\f$ of mixture components.

Python prototype (for reference only):

trainE(samples, means0[, covs0[, weights0[, logLikelihoods[, labels[, probs]]]]]) -> retval, logLikelihoods, labels, probs
Link to this function

trainE(self, samples, means0, opts)

View Source
@spec trainE(
  t(),
  Evision.Mat.maybe_mat_in(),
  Evision.Mat.maybe_mat_in(),
  [covs0: term(), weights0: term()] | nil
) ::
  {Evision.Mat.t(), Evision.Mat.t(), Evision.Mat.t()}
  | false
  | {:error, String.t()}

Estimate the Gaussian mixture parameters from a samples set.

Positional Arguments
  • self: Evision.ML.EM.t()

  • samples: Evision.Mat.

    Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.

  • means0: Evision.Mat.

    Initial means \f$a_k\f$ of mixture components. It is a one-channel matrix of \f$nclusters \times dims\f$ size. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.

Keyword Arguments
  • covs0: Evision.Mat.

    The vector of initial covariance matrices \f$S_k\f$ of mixture components. Each of covariance matrices is a one-channel matrix of \f$dims \times dims\f$ size. If the matrices do not have CV_64F type they will be converted to the inner matrices of such type for the further computing.

  • weights0: Evision.Mat.

    Initial weights \f$\pi_k\f$ of mixture components. It should be a one-channel floating-point matrix with \f$1 \times nclusters\f$ or \f$nclusters \times 1\f$ size.

Return
  • retval: bool

  • logLikelihoods: Evision.Mat.t().

    The optional output matrix that contains a likelihood logarithm value for each sample. It has \f$nsamples \times 1\f$ size and CV_64FC1 type.

  • labels: Evision.Mat.t().

    The optional output "class label" for each sample: \f$\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\f$ (indices of the most probable mixture component for each sample). It has \f$nsamples \times 1\f$ size and CV_32SC1 type.

  • probs: Evision.Mat.t().

    The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \f$nsamples \times nclusters\f$ size and CV_64FC1 type.

This variation starts with Expectation step. You need to provide initial means \f$a\_k\f$ of mixture components. Optionally you can pass initial weights \f$\pi\_k\f$ and covariance matrices \f$S\_k\f$ of mixture components.

Python prototype (for reference only):

trainE(samples, means0[, covs0[, weights0[, logLikelihoods[, labels[, probs]]]]]) -> retval, logLikelihoods, labels, probs
@spec trainEM(Keyword.t()) :: any() | {:error, String.t()}
@spec trainEM(t(), Evision.Mat.maybe_mat_in()) ::
  {Evision.Mat.t(), Evision.Mat.t(), Evision.Mat.t()}
  | false
  | {:error, String.t()}

Estimate the Gaussian mixture parameters from a samples set.

Positional Arguments
  • self: Evision.ML.EM.t()

  • samples: Evision.Mat.

    Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.

Return
  • retval: bool

  • logLikelihoods: Evision.Mat.t().

    The optional output matrix that contains a likelihood logarithm value for each sample. It has \f$nsamples \times 1\f$ size and CV_64FC1 type.

  • labels: Evision.Mat.t().

    The optional output "class label" for each sample: \f$\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\f$ (indices of the most probable mixture component for each sample). It has \f$nsamples \times 1\f$ size and CV_32SC1 type.

  • probs: Evision.Mat.t().

    The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \f$nsamples \times nclusters\f$ size and CV_64FC1 type.

This variation starts with Expectation step. Initial values of the model parameters will be estimated by the k-means algorithm. Unlike many of the ML models, %EM is an unsupervised learning algorithm and it does not take responses (class labels or function values) as input. Instead, it computes the Maximum Likelihood Estimate of the Gaussian mixture parameters from an input sample set, stores all the parameters inside the structure: \f$p\_{i,k}\f$ in probs, \f$a\_k\f$ in means , \f$S\_k\f$ in covs[k], \f$\pi\_k\f$ in weights , and optionally computes the output "class label" for each sample: \f$\texttt{labels}\_i=\texttt{arg max}\_k(p\_{i,k}), i=1..N\f$ (indices of the most probable mixture component for each sample). The trained model can be used further for prediction, just like any other classifier. The trained model is similar to the NormalBayesClassifier.

Python prototype (for reference only):

trainEM(samples[, logLikelihoods[, labels[, probs]]]) -> retval, logLikelihoods, labels, probs
Link to this function

trainEM(self, samples, opts)

View Source
@spec trainEM(t(), Evision.Mat.maybe_mat_in(), [{atom(), term()}, ...] | nil) ::
  {Evision.Mat.t(), Evision.Mat.t(), Evision.Mat.t()}
  | false
  | {:error, String.t()}

Estimate the Gaussian mixture parameters from a samples set.

Positional Arguments
  • self: Evision.ML.EM.t()

  • samples: Evision.Mat.

    Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.

Return
  • retval: bool

  • logLikelihoods: Evision.Mat.t().

    The optional output matrix that contains a likelihood logarithm value for each sample. It has \f$nsamples \times 1\f$ size and CV_64FC1 type.

  • labels: Evision.Mat.t().

    The optional output "class label" for each sample: \f$\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\f$ (indices of the most probable mixture component for each sample). It has \f$nsamples \times 1\f$ size and CV_32SC1 type.

  • probs: Evision.Mat.t().

    The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \f$nsamples \times nclusters\f$ size and CV_64FC1 type.

This variation starts with Expectation step. Initial values of the model parameters will be estimated by the k-means algorithm. Unlike many of the ML models, %EM is an unsupervised learning algorithm and it does not take responses (class labels or function values) as input. Instead, it computes the Maximum Likelihood Estimate of the Gaussian mixture parameters from an input sample set, stores all the parameters inside the structure: \f$p\_{i,k}\f$ in probs, \f$a\_k\f$ in means , \f$S\_k\f$ in covs[k], \f$\pi\_k\f$ in weights , and optionally computes the output "class label" for each sample: \f$\texttt{labels}\_i=\texttt{arg max}\_k(p\_{i,k}), i=1..N\f$ (indices of the most probable mixture component for each sample). The trained model can be used further for prediction, just like any other classifier. The trained model is similar to the NormalBayesClassifier.

Python prototype (for reference only):

trainEM(samples[, logLikelihoods[, labels[, probs]]]) -> retval, logLikelihoods, labels, probs
@spec trainM(Keyword.t()) :: any() | {:error, String.t()}
Link to this function

trainM(self, samples, probs0)

View Source
@spec trainM(t(), Evision.Mat.maybe_mat_in(), Evision.Mat.maybe_mat_in()) ::
  {Evision.Mat.t(), Evision.Mat.t(), Evision.Mat.t()}
  | false
  | {:error, String.t()}

Estimate the Gaussian mixture parameters from a samples set.

Positional Arguments
  • self: Evision.ML.EM.t()

  • samples: Evision.Mat.

    Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.

  • probs0: Evision.Mat.

    the probabilities

Return
  • retval: bool

  • logLikelihoods: Evision.Mat.t().

    The optional output matrix that contains a likelihood logarithm value for each sample. It has \f$nsamples \times 1\f$ size and CV_64FC1 type.

  • labels: Evision.Mat.t().

    The optional output "class label" for each sample: \f$\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\f$ (indices of the most probable mixture component for each sample). It has \f$nsamples \times 1\f$ size and CV_32SC1 type.

  • probs: Evision.Mat.t().

    The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \f$nsamples \times nclusters\f$ size and CV_64FC1 type.

This variation starts with Maximization step. You need to provide initial probabilities \f$p\_{i,k}\f$ to use this option.

Python prototype (for reference only):

trainM(samples, probs0[, logLikelihoods[, labels[, probs]]]) -> retval, logLikelihoods, labels, probs
Link to this function

trainM(self, samples, probs0, opts)

View Source
@spec trainM(
  t(),
  Evision.Mat.maybe_mat_in(),
  Evision.Mat.maybe_mat_in(),
  [{atom(), term()}, ...] | nil
) ::
  {Evision.Mat.t(), Evision.Mat.t(), Evision.Mat.t()}
  | false
  | {:error, String.t()}

Estimate the Gaussian mixture parameters from a samples set.

Positional Arguments
  • self: Evision.ML.EM.t()

  • samples: Evision.Mat.

    Samples from which the Gaussian mixture model will be estimated. It should be a one-channel matrix, each row of which is a sample. If the matrix does not have CV_64F type it will be converted to the inner matrix of such type for the further computing.

  • probs0: Evision.Mat.

    the probabilities

Return
  • retval: bool

  • logLikelihoods: Evision.Mat.t().

    The optional output matrix that contains a likelihood logarithm value for each sample. It has \f$nsamples \times 1\f$ size and CV_64FC1 type.

  • labels: Evision.Mat.t().

    The optional output "class label" for each sample: \f$\texttt{labels}_i=\texttt{arg max}_k(p_{i,k}), i=1..N\f$ (indices of the most probable mixture component for each sample). It has \f$nsamples \times 1\f$ size and CV_32SC1 type.

  • probs: Evision.Mat.t().

    The optional output matrix that contains posterior probabilities of each Gaussian mixture component given the each sample. It has \f$nsamples \times nclusters\f$ size and CV_64FC1 type.

This variation starts with Maximization step. You need to provide initial probabilities \f$p\_{i,k}\f$ to use this option.

Python prototype (for reference only):

trainM(samples, probs0[, logLikelihoods[, labels[, probs]]]) -> retval, logLikelihoods, labels, probs
@spec write(Keyword.t()) :: any() | {:error, String.t()}
@spec write(t(), Evision.FileStorage.t()) :: t() | {:error, String.t()}

Stores algorithm parameters in a file storage

Positional Arguments

Python prototype (for reference only):

write(fs) -> None
@spec write(t(), Evision.FileStorage.t(), binary()) :: t() | {:error, String.t()}

write

Positional Arguments

Has overloading in C++

Python prototype (for reference only):

write(fs, name) -> None