Object

com.salesforce.op.stages.impl.classification

BinaryClassificationModelSelector

Related Doc: package classification

Permalink

object BinaryClassificationModelSelector extends ModelSelectorFactory with Product with Serializable

A factory for Binary Classification Model Selector

Linear Supertypes
Serializable, Serializable, Product, Equals, ModelSelectorFactory, AnyRef, Any
Ordering
  1. Alphabetic
  2. By Inheritance
Inherited
  1. BinaryClassificationModelSelector
  2. Serializable
  3. Serializable
  4. Product
  5. Equals
  6. ModelSelectorFactory
  7. AnyRef
  8. Any
  1. Hide All
  2. Show All
Visibility
  1. Public
  2. All

Value Members

  1. final def !=(arg0: Any): Boolean

    Permalink
    Definition Classes
    AnyRef → Any
  2. final def ##(): Int

    Permalink
    Definition Classes
    AnyRef → Any
  3. final def ==(arg0: Any): Boolean

    Permalink
    Definition Classes
    AnyRef → Any
  4. def apply(): ModelSelector[ModelType, EstimatorType]

    Permalink

    Creates a new Binary Classification Model Selector with a Cross Validation

  5. final def asInstanceOf[T0]: T0

    Permalink
    Definition Classes
    Any
  6. def clone(): AnyRef

    Permalink
    Attributes
    protected[java.lang]
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  7. def defaultModelsAndParams: Seq[(EstimatorType, Array[ParamMap])]

    Permalink

    Default models and parameters

    Default models and parameters

    returns

    defaults for problem type

    Attributes
    protected
    Definition Classes
    BinaryClassificationModelSelector → ModelSelectorFactory
  8. final def eq(arg0: AnyRef): Boolean

    Permalink
    Definition Classes
    AnyRef
  9. def equals(arg0: Any): Boolean

    Permalink
    Definition Classes
    AnyRef → Any
  10. def finalize(): Unit

    Permalink
    Attributes
    protected[java.lang]
    Definition Classes
    AnyRef
    Annotations
    @throws( classOf[java.lang.Throwable] )
  11. final def getClass(): Class[_]

    Permalink
    Definition Classes
    AnyRef → Any
  12. final def isInstanceOf[T0]: Boolean

    Permalink
    Definition Classes
    Any
  13. final def ne(arg0: AnyRef): Boolean

    Permalink
    Definition Classes
    AnyRef
  14. final def notify(): Unit

    Permalink
    Definition Classes
    AnyRef
  15. final def notifyAll(): Unit

    Permalink
    Definition Classes
    AnyRef
  16. def selector(validator: OpValidator[ModelType, EstimatorType], splitter: Option[Splitter], trainTestEvaluators: Seq[OpEvaluatorBase[_ <: EvaluationMetrics]], modelTypesToUse: Seq[ModelsToTry], modelsAndParameters: Seq[(EstimatorType, Array[ParamMap])]): ModelSelector[ModelType, EstimatorType]

    Permalink

    Create the model selector for specified interface

    Create the model selector for specified interface

    validator

    training split of cross validator

    splitter

    data prep class

    trainTestEvaluators

    evaluation to do on data

    modelTypesToUse

    list of models to use

    modelsAndParameters

    sequence of models and parameters to explore

    returns

    model selector with these settings

    Attributes
    protected
    Definition Classes
    ModelSelectorFactory
  17. final def synchronized[T0](arg0: ⇒ T0): T0

    Permalink
    Definition Classes
    AnyRef
  18. final def wait(): Unit

    Permalink
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  19. final def wait(arg0: Long, arg1: Int): Unit

    Permalink
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  20. final def wait(arg0: Long): Unit

    Permalink
    Definition Classes
    AnyRef
    Annotations
    @throws( ... )
  21. def withCrossValidation(splitter: Option[Splitter] = Option(DataSplitter()), numFolds: Int = ValidatorParamDefaults.NumFolds, validationMetric: OpBinaryClassificationEvaluatorBase[_] = ..., trainTestEvaluators: Seq[OpBinaryClassificationEvaluatorBase[_ <: EvaluationMetrics]] = Seq.empty, seed: Long = ValidatorParamDefaults.Seed, stratify: Boolean = ValidatorParamDefaults.Stratify, parallelism: Int = ValidatorParamDefaults.Parallelism, modelTypesToUse: Seq[BinaryClassificationModelsToTry] = modelNames, modelsAndParameters: Seq[(EstimatorType, Array[ParamMap])] = Seq.empty): ModelSelector[ModelType, EstimatorType]

    Permalink

    Creates a new Binary Classification Model Selector with a Cross Validation

    Creates a new Binary Classification Model Selector with a Cross Validation

    splitter

    instance that will balance and split the data

    numFolds

    number of folds for cross validation (>= 2)

    validationMetric

    metric name in evaluation: AuROC or AuPR

    trainTestEvaluators

    List of evaluators applied on training + holdout data for evaluation. Default is empty and default evaluator is added to this list (here Evaluators.BinaryClassification)

    seed

    random seed

    stratify

    whether or not stratify cross validation. Caution : setting that param to true might impact the runtime

    parallelism

    level of parallelism used to schedule a number of models to be trained/evaluated so that the jobs can be run concurrently

    modelTypesToUse

    list of model types to run grid search on must from supported types in BinaryClassificationModelsToTry (OpLogisticRegression, OpRandomForestClassifier, OpGBTClassifier, OpLinearSVC, OpDecisionTreeClassifier, OpNaiveBayes)

    modelsAndParameters

    pass in an explicit list pairs of estimators and the accompanying hyperparameters to for model selection Seq[(EstimatorType, Array[ParamMap])] where Estimator type must be an Estimator that takes in a label (RealNN) and features (OPVector) and returns a prediction (Prediction)

    returns

    Classification Model Selector with a Cross Validation

  22. def withTrainValidationSplit(splitter: Option[Splitter] = Option(DataSplitter()), trainRatio: Double = ValidatorParamDefaults.TrainRatio, validationMetric: OpBinaryClassificationEvaluatorBase[_] = ..., trainTestEvaluators: Seq[OpBinaryClassificationEvaluatorBase[_ <: EvaluationMetrics]] = Seq.empty, seed: Long = ValidatorParamDefaults.Seed, stratify: Boolean = ValidatorParamDefaults.Stratify, parallelism: Int = ValidatorParamDefaults.Parallelism, modelTypesToUse: Seq[BinaryClassificationModelsToTry] = modelNames, modelsAndParameters: Seq[(EstimatorType, Array[ParamMap])] = Seq.empty): ModelSelector[ModelType, EstimatorType]

    Permalink

    Creates a new Binary Classification Model Selector with a Train Validation Split

    Creates a new Binary Classification Model Selector with a Train Validation Split

    splitter

    instance that will balance and split the data

    trainRatio

    ratio between training set and validation set (>= 0 && <= 1)

    validationMetric

    metric name in evaluation: AuROC or AuPR

    trainTestEvaluators

    List of evaluators applied on training + holdout data for evaluation. Default is empty and default evaluator is added to this list (here Evaluators.BinaryClassification)

    seed

    random seed

    stratify

    whether or not stratify train validation split. Caution : setting that param to true might impact the runtime

    parallelism

    level of parallelism used to schedule a number of models to be trained/evaluated so that the jobs can be run concurrently

    modelTypesToUse

    list of model types to run grid search on must from supported types in BinaryClassificationModelsToTry (OpLogisticRegression, OpRandomForestClassifier, OpGBTClassifier, OpLinearSVC, OpDecisionTreeClassifier, OpNaiveBayes)

    modelsAndParameters

    pass in an explicit list pairs of estimators and the accompanying hyperparameters to for model selection Seq[(EstimatorType, Array[ParamMap])] where Estimator type must be an Estimator that takes in a label (RealNN) and features (OPVector) and returns a prediction (Prediction)

    returns

    Classification Model Selector with a Train Validation Split

Inherited from Serializable

Inherited from Serializable

Inherited from Product

Inherited from Equals

Inherited from ModelSelectorFactory

Inherited from AnyRef

Inherited from Any

Ungrouped