Global

Type Definitions

gmmResults

Results of the filtering process.

Type:
  • Object
Properties:
Name Type Attributes Description
likeliest String

The likeliest model's label.
likeliestIndex Number

The likeliest model's index
likelihoods Array.number

The array of all models' smoothed normalized likelihoods.
outputValues Array.number <nullable>

If the model was trained with regression, the estimated float vector output.
outputCovariance Array.number <nullable>

If the model was trained with regression, the output covariance matrix.
Source:

gmmResultsCallback(err, res)

Callback handling estimation results.

Parameters:
Name Type Description
err String

Description of a potential error.
res gmmResults

Object holding the estimation results.
Source:

hhmmResults

Results of the filtering process.

Type:
  • Object
Properties:
Name Type Attributes Description
likeliest String

The likeliest model's label.
likeliestIndex Number

The likeliest model's index
likelihoods Array.number

The array of all models' smoothed normalized likelihoods.
timeProgressions Array.number

The array of all models' normalized time progressions.
alphas Array.Array.number

The array of all models' states likelihoods array.
outputValues Array.number <nullable>

If the model was trained with regression, the estimated float vector output.
outputCovariance Array.number <nullable>

If the model was trained with regression, the output covariance matrix.
Source:

hhmmResultsCallback(err, res)

Callback handling estimation results.

Parameters:
Name Type Description
err string

Description of a potential error.
res hhmmResults

Object holding the estimation results.
Source:

xmmGmmModel

A valid XMM GMM model

Type:
  • Object
Properties:
Name Type Description
TODO String

LIST REAL GMM MODEL PROPERTIES HERE
Source:

xmmHhmmModel

A valid XMM Hierarchical HMM model

Type:
  • Object
Properties:
Name Type Description
TODO String

LIST REAL HHMM MODEL PROPERTIES HERE
Source:

xmmPhrase

A valid XMM phrase, ready to be processed by the XMM library.

Type:
  • Object
Properties:
Name Type Description
bimodal Boolean

Indicates wether phrase data should be considered bimodal. If true, the dimension_input property will be taken into account.
dimension Number

Size of a phrase's vector element.
dimension_input Number

Size of the part of an input vector element that should be used for training. This implies that the rest of the vector (of size dimension - dimension_input) will be used for regression. Only taken into account if bimodal is true.
column_names Array.String

Array of string identifiers describing each scalar of the phrase's vector elements. Typically of size dimension.
label String

The string identifier of the class the phrase belongs to.
data Array.Number

The phrase's data, containing all the vectors flattened into a single one. Only taken into account if bimodal is false.
data_input Array.Number

The phrase's data which will be used for training, flattened into a single vector. Only taken into account if bimodal is true.
data_output Array.Number

The phrase's data which will be used for regression, flattened into a single vector. Only taken into account if bimodal is true.
length Number

The length of the phrase, e.g. one of the following :

    • data.length / dimension
      data_input.length / dimension_input
      data_output.length / dimension_output
    		•
    Source:

    xmmPhraseConfig

    XMM phrase configuration object.

    Type:
    • Object
    Properties:
    Name Type Description
    bimodal Boolean

    Indicates wether phrase data should be considered bimodal. If true, the dimension_input property will be taken into account.
    dimension Number

    Size of a phrase's vector element.
    dimensionInput Number

    Size of the part of an input vector element that should be used for training. This implies that the rest of the vector (of size dimension - dimension_input) will be used for regression. Only taken into account if bimodal is true.
    column_names Array.String

    Array of string identifiers describing each scalar of the phrase's vector elements. Typically of size dimension.
    label String

    The string identifier of the class the phrase belongs to.
    Source:

    xmmTrainingSet

    A valid XMM training set, ready to be processed by the XMM library.

    Type:
    • Object
    Properties:
    Name Type Description
    bimodal Boolean

    Indicates wether the set's phrases data should be considered bimodal. If true, the dimension_input property will be taken into account.
    dimension Number

    Size of a vector element of the set's phrases.
    dimension_input Number

    Size of the part of an input vector element that should be used for training. This implies that the rest of the vector (of size dimension - dimension_input) will be used for regression. Only taken into account if bimodal is true.
    column_names Array.String

    Array of string identifiers describing each scalar of a phrase's vector elements. Typically of size dimension.
    phrases Array.xmmPhrase

    Array of valid XMM phrases containing an extra "index" field.
    Source: