public static final class GradientBoosting.LossFunctionType extends Enum
Modifier and Type | Field and Description |
---|---|
static GradientBoosting.LossFunctionType |
ADABOOST
The loss criteria is the AdaBoost.M1 criterion.
|
static GradientBoosting.LossFunctionType |
BERNOULLI
The loss criteria is the binomial or Bernoulli negative
log-likelihood, or deviance.
|
static GradientBoosting.LossFunctionType |
HUBER_M
The loss criteria is the Huber-M weighted squared error and absolute
deviation error with parameter .
|
static GradientBoosting.LossFunctionType |
LEAST_ABSOLUTE_DEVIATION
The loss criteria is least absolute deviation error.
|
static GradientBoosting.LossFunctionType |
LEAST_SQUARES
The loss criteria is least squared error.
|
static GradientBoosting.LossFunctionType |
MULTINOMIAL_DEVIANCE
The loss criteria is the (K-class) multinomial negative
log-likelihood, or multinomial deviance.
|
Modifier and Type | Method and Description |
---|---|
static GradientBoosting.LossFunctionType |
valueOf(String name) |
static GradientBoosting.LossFunctionType[] |
values() |
public static final GradientBoosting.LossFunctionType ADABOOST
public static final GradientBoosting.LossFunctionType BERNOULLI
public static final GradientBoosting.LossFunctionType HUBER_M
where
And where is the -empirical quantile of the residuals,
public static final GradientBoosting.LossFunctionType LEAST_ABSOLUTE_DEVIATION
public static final GradientBoosting.LossFunctionType LEAST_SQUARES
public static final GradientBoosting.LossFunctionType MULTINOMIAL_DEVIANCE
where
public static GradientBoosting.LossFunctionType valueOf(String name)
public static GradientBoosting.LossFunctionType[] values()
Copyright © 1970-2015 Rogue Wave Software
Built October 13 2015.