DataPredict

API Reference - Others - OneVsAll

Allows binary classification models (such as LogisticRegression) be merged together to form multi-class models.

Constructors

new()

Create new model object. If any of the arguments are nil, default argument values for that argument will be used.

OneVsAll.new(maxNumberOfIterations: integer, useNegativeOneBinaryLabel: boolean): OneVsAllObject

Parameters:

• maxNumberOfIterations: How many times should the model needed to be trained.

• useNegativeOneBinaryLabel: Set whether or not if the negative labels uses -1 instead of 0.

Returns:

• OneVsAllObject: The generated OneVsAll object.

Functions

setParameters()

Set model’s parameters. When any of the arguments are nil, previous argument values for that argument will be used.

OneVsAll:setParameters(maxNumberOfIterations: integer, useNegativeOneBinaryLabel: boolean)

Parameters:

• maxNumberOfIterations: How many times should the model needed to be trained.

• useNegativeOneBinaryLabel: Set whether or not if the negative labels uses -1 instead of 0.

setModels()

Sets the model and number of classes to be used by the OneVsAll object. Leaving it empty will clear the model.

OneVsAll:setModels(modelName: string, numberOfClasses: integer)

Parameters:

• modelName: The full name of the model to be used in OneVsAll object.

• numberOfClasses: The number of models to be generated based on number of classes.

setOptimizer()

Sets the optimizer and its parameters. Leaving it empty will clear the optimizer.

OneVsAll:setOptimizer(optimizerName: string, ...)

Parameters:

• optimizerName: The full name of the optimizer to be used in OneVsAll object.

• …: The parameters to be provided to the optimizer.

setRegularizer()

Sets the regularizer and its parameters. Leaving it empty will clear the optimizer.

OneVsAll:setRegularizer(lambda: number, regularizationMode: string, hasBias: boolean)

Parameters:

• lambda: Regularization factor. Recommended values are between 0 to 1.

• regularisationMode: The mode which regularization will be used. Currently available ones are “L1” (or “Lasso”), “L2” (or “Ridge”) and “L1+L2” (or “ElasticNet”).

• hasBias: Set whether or not the regularization has bias.

setModelsSettings()

OneVsAll:setModelsSettings(...: any)

Parameters:

• …: The parameters to be set to all models stored in this OneVsAll object.

train()

Train the model.

NeuralNetwork:train(featureMatrix: Matrix, labelVector / labelMatrix: Matrix): number[]

Parameters:

• featureMatrix: Matrix containing all data.

• labelVector / labelMatrix: A (n x 1) / (n x o) matrix containing values related to featureMatrix. When using the label matrix, the number of columns must be equal to number of classes.

Returns:

• costArray: An array containing cost values.

predict()

Predict the values for given data.

OneVsAll:predict(featureMatrix: Matrix): Matrix, Matrix

Parameters:

• featureMatrix: Matrix containing all data.

• returnOriginalOutput: Set whether or not to return predicted matrix instead of value with highest probability.

Returns:

• predictedVector: A vector that is predicted by the model.

• highestValueVector: A vector that contains the predicted values in predictedVector.

getClassesList()

OneVsAll:getClassesList(): []

Returns:

• ClassesList: A list of classes. The index of the list relates to which model belong to. For example, {3, 1} means that the output for 3 is at first model, and the output for 1 is at second model.

setClassesList()

OneVsAll:setClassesList(ClassesList: [])

Parameters:

• ClassesList: A list of classes. The index of the list relates to which model belong to. For example, {3, 1} means that the output for 3 is at first model, and the output for 1 is at second model.

getModelParametersArray()

Gets the model parameters from the base model.

OneVsAll:getModelParametersArray(doNotDeepCopy: boolean): ModelParameters []

Parameters

• doNotDeepCopy: Set whether or not to deep copy the model parameters.

Returns

• ModelParameters: An array containing model parameters (matrix/table) fetched from each model. The index of the array determines which model it belongs to.

setModelParametersArray()

Set the model parameters to the base model.

OneVsAll:setModelParameters(ModelParametersArray: ModelParameters[], doNotDeepCopy: boolean)

Parameters

• ModelParametersArray: A table containing model parameters (matrix/table) to be given to be given to each model stored in OneVsAll object. The position of the parameters determines which model it belongs to.

• doNotDeepCopy: Set whether or not to deep copy the model parameters.

clearModelParameters()

Clears the model parameters stored inside the models.

OneVsAll:clearModelParameters()

setNumberOfIterationsToCheckIfConverged()

Set the number of iterations needed to confirm convergence for each model.

OneVsAll:setNumberOfIterationsToCheckIfConverged(numberOfIterations: number)

Parameters

• numberOfIterations: The number of iterations for confirming convergence.

setNumberOfIterationsToCheckIfConvergedForOneVsAll()

Set the number of iterations needed to confirm convergence.

OneVsAll:setNumberOfIterationsToCheckIfConvergedForOneVsAll(numberOfIterations: number)

Parameters

• numberOfIterations: The number of iterations for confirming convergence.

setTargetCost()

Set the upper bound and lower bounds of the target cost for each model.

OneVsAll:setTargetCost(upperBound: number, lowerBound: number)

Parameters

• upperBound: The upper bound of target cost.

• lowerBound: The lower bound of target cost.

setTargetTotalCost()

Set the upper bound and lower bounds of the target cost.

OneVsAll:setTargetTotalCost(upperBound: number, lowerBound: number)

Parameters

• upperBound: The upper bound of target cost.

• lowerBound: The lower bound of target cost.

setAutoResetOptimizers()

Set if the optimizer resets at the end of iterations.

OneVsAll:setAutoResetOptimizers(option: boolean)

Parameters:

• option: A boolean value that specifies if optimizers resets at the end of iterations.

setPrintOutput()

Set if the OneVsAll object prints output.

OneVsAll:setPrintOutput(option: boolean)

Parameters:

• option: A boolean value that specifies if the output is printed.