org.apache.spark.ml.classification

Class LogisticAggregator

Object

org.apache.spark.ml.classification.LogisticAggregator

All Implemented Interfaces:

java.io.Serializable

public class LogisticAggregator
extends Object
implements scala.Serializable

LogisticAggregator computes the gradient and loss for binary logistic loss function, as used in binary classification for instances in sparse or dense vector in an online fashion.

Note that multinomial logistic loss is not supported yet!

Two LogisticAggregator can be merged together to have a summary of loss and gradient of the corresponding joint dataset.

param: numClasses the number of possible outcomes for k classes classification problem in Multinomial Logistic Regression. param: fitIntercept Whether to fit an intercept term.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
LogisticAggregator(int numFeatures, int numClasses, boolean fitIntercept)

Method Summary

Methods

Modifier and Type	Method and Description
LogisticAggregator	add(org.apache.spark.ml.feature.Instance instance, Vector coefficients, double[] featuresStd) Add a new training instance to this LogisticAggregator, and update the loss and gradient of the objective function.
Vector	gradient()
double	loss()
LogisticAggregator	merge(LogisticAggregator other) Merge another LogisticAggregator, and update the loss and gradient of the objective function.

Methods inherited from class Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

LogisticAggregator

public LogisticAggregator(int numFeatures,
                  int numClasses,
                  boolean fitIntercept)

Method Detail

add

public LogisticAggregator add(org.apache.spark.ml.feature.Instance instance,
                     Vector coefficients,
                     double[] featuresStd)

Add a new training instance to this LogisticAggregator, and update the loss and gradient of the objective function.

Parameters:

instance - The instance of data point to be added.

coefficients - The coefficients corresponding to the features.

featuresStd - The standard deviation values of the features.

Returns:

This LogisticAggregator object.

merge

public LogisticAggregator merge(LogisticAggregator other)

Merge another LogisticAggregator, and update the loss and gradient of the objective function. (Note that it's in place merging; as a result, this object will be modified.)

Parameters:

other - The other LogisticAggregator to be merged.

Returns:

This LogisticAggregator object.

loss

public double loss()

gradient

public Vector gradient()