smile.regression

Class RandomForest.Trainer

java.lang.Object

smile.regression.RegressionTrainer<double[]>

smile.regression.RandomForest.Trainer

Enclosing class:

RandomForest

public static class RandomForest.Trainer
extends RegressionTrainer<double[]>

Trainer for random forest.

Constructor Summary

Constructors

Constructor and Description
RandomForest.Trainer(Attribute[] attributes, int T) Constructor.
RandomForest.Trainer(int T) Constructor.

Method Summary

Methods

Modifier and Type	Method and Description
void	setMinimumNodeSize(int S) Sets the minimum size of leaf nodes.
void	setNumRandomFeatures(int M) Sets the number of random selected features for splitting.
void	setNumTrees(int T) Sets the number of trees in the random forest.
RandomForest	train(double[][] x, double[] y) Learns a regression model with given training data.

Methods inherited from class smile.regression.RegressionTrainer

setAttributes

Methods inherited from class java.lang.Object

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

Constructor Detail

RandomForest.Trainer

public RandomForest.Trainer(int T)

Constructor.

Parameters:

T - the number of trees.

RandomForest.Trainer

public RandomForest.Trainer(Attribute[] attributes,
                    int T)

Constructor.

Parameters:

attributes - the attributes of independent variable.

T - the number of trees.

Method Detail

setNumTrees

public void setNumTrees(int T)

Sets the number of trees in the random forest.

Parameters:

T - the number of trees.

setNumRandomFeatures

public void setNumRandomFeatures(int M)

Sets the number of random selected features for splitting.

Parameters:

M - the number of random selected features to be used to determine the decision at a node of the tree. floor(sqrt(dim)) seems to give generally good performance, where dim is the number of variables.

setMinimumNodeSize

public void setMinimumNodeSize(int S)

Sets the minimum size of leaf nodes.

Parameters:

S - the number of instances in a node below which the tree will not split, setting S = 5 generally gives good results.

train

public RandomForest train(double[][] x,
                 double[] y)

Description copied from class: RegressionTrainer

Learns a regression model with given training data.

Specified by:

train in class RegressionTrainer<double[]>

Parameters:

x - the training instances.

y - the training response values.

Returns:

a trained regression model.