ml

Type Members

1. case class ClusterInfo(name: String, centroid: Seq[Double]) extends Product with Serializable

   

   Representation of a single cluster.

   Representation of a single cluster.

   We use java.lang.Double for the type of the numeric values, because the scala Double type information is lost by scala/Gson and the deserialization fails badly for edge cases (e.g. Double.NaN).

   name

   Name used to distinguish this cluster from others in the same model.

   centroid

   A vector representation of the cluster in orthogonal coordinates.

2. case class KMeansModel(clusters: Seq[ClusterInfo], inputFeatures: Seq[ColumnDef], identifier: String = "") extends ClusteringRowModel with PFAConvertible with Product with Serializable

   

   A model representing results of the K-Means clustering algorithm.

   A model representing results of the K-Means clustering algorithm. Each cluster is a vector in a fixed dimensional space, and the transformer assigns the input row to its closest cluster using the Euclidean (L2) distance.

   The clusters should all be of the same dimension, which should be equal to the number of input features.

   clusters

   The clusters of the model. These should have distinct names.

   inputFeatures

   A seq of numeric feature descriptions describing the input to the model. * @param identifier Used to identify this model when in a collection of models. Should be simple characters, so it can be used in a feature name.

   Annotations

   @SerialVersionUID()

3. class KMeansSQLTransformer extends ClusteringSQLTransformer

   

4. case class KMeansTransformer(model: KMeansModel) extends ClusteringTransformer with Product with Serializable

   

5. case class LinearRegressionModel(coefficients: Seq[Double], inputFeatures: Seq[ColumnDef], intercept: Double = 0, dependentFeatureName: String = "", identifier: String = "") extends RegressionRowModel with PFAConvertible with Product with Serializable

   

   Representation of the classical linear regression model y = intercept + a_0 x_0 + a_1 x_1 + ...

   Representation of the classical linear regression model y = intercept + a_0 x_0 + a_1 x_1 + ... + a_n x_n.

   The length of the coefficients must match the length of the inputFeatures.

   We use java.lang.Double for the type of the coefficients, because the scala Double type information is lost by scala/Gson and the deserialization fails badly for edge cases (e.g. Double.NaN).

   coefficients

   Vector of coefficients of the model. Must match the length of inputFeatures.

   inputFeatures

   Description of the (numeric) input features. Must match the length of coefficients.

   intercept

   Intercept of the model (defaults to 0).

   dependentFeatureName

   Name used to identify the dependent feature in an evaluation dataset.

   identifier

   Used to identify this model when in a collection of models. Should be simple characters, so it can be used in a feature name.

   Annotations

   @SerialVersionUID()

6. class LinearRegressionTransformer extends RegressionTransformer

   

   Applies a Linear Regression model specified by the coefficients and intercept to a row of numeric data.

   Applies a Linear Regression model specified by the coefficients and intercept to a row of numeric data.

   Note that in the input row is wrapped in CastedDoubleSeq, so the input elements must be castable as java.lang.Number.

7. case class LogisticRegressionSQLTransformer(model: MultiLogisticRegressionModel, sqlGenerator: SQLGenerator) extends ClassificationSQLTransformer with Product with Serializable

   

8. case class LogisticRegressionTransformer(model: MultiLogisticRegressionModel) extends ClassificationTransformer with Product with Serializable

   

9. case class MultiLogisticRegressionModel(singleLORs: Seq[SingleLogisticRegression], baseValue: String, dependentFeatureName: String, inputFeatures: Seq[ColumnDef], identifier: String = "") extends ClassificationRowModel with PFAConvertible with Product with Serializable

   
   Annotations

   @SerialVersionUID()

10. case class SingleLogisticRegression(dependentValue: String, coefficients: Seq[Double], bias: Double = 0) extends Product with Serializable

    

    Represents a SingleLogisticRegression to be used as one of several in a MultiLogisticRegressionModel.

    Represents a SingleLogisticRegression to be used as one of several in a MultiLogisticRegressionModel.

    We use java.lang.Double for the type of the numeric values, because the scala Double type information is lost by scala/Gson and the deserialization fails badly for edge cases (e.g. Double.NaN).

    dependentValue

    The dependent value that the coefficients correspond to.

    coefficients

    The coefficients for the single Logistic Regression model.

    bias

    The constant term, that is added to the dot product of the feature coefficient vectors.