smile.validation

Class Bootstrap

java.lang.Object

smile.validation.Bootstrap

public class Bootstrap
extends java.lang.Object

The bootstrap is a general tool for assessing statistical accuracy. The basic idea is to randomly draw datasets with replacement from the training data, each samples the same size as the original training set. This is done many times (say k = 100), producing k bootstrap datasets. Then we refit the model to each of the bootstrap datasets and examine the behavior of the fits over the k replications.

Field Summary

Fields

Modifier and Type	Field and Description
int	k The number of rounds of cross validation.
int[][]	test The index of testing instances.
int[][]	train The index of training instances.

Constructor Summary

Constructors

Constructor and Description
Bootstrap(int n, int k) Constructor.

Method Summary

Modifier and Type	Method and Description
double[]	classification(smile.data.formula.Formula formula, smile.data.DataFrame data, java.util.function.BiFunction<smile.data.formula.Formula,smile.data.DataFrame,DataFrameClassifier> trainer) Runs cross validation tests.
static double[]	classification(int k, smile.data.formula.Formula formula, smile.data.DataFrame data, java.util.function.BiFunction<smile.data.formula.Formula,smile.data.DataFrame,DataFrameClassifier> trainer) Runs cross validation tests.
static <T> double[]	classification(int k, T[] x, int[] y, java.util.function.BiFunction<T[],int[],Classifier<T>> trainer) Runs cross validation tests.
<T> double[]	classification(T[] x, int[] y, java.util.function.BiFunction<T[],int[],Classifier<T>> trainer) Runs cross validation tests.
double[]	regression(smile.data.formula.Formula formula, smile.data.DataFrame data, java.util.function.BiFunction<smile.data.formula.Formula,smile.data.DataFrame,DataFrameRegression> trainer) Runs bootstrap tests.
static double[]	regression(int k, smile.data.formula.Formula formula, smile.data.DataFrame data, java.util.function.BiFunction<smile.data.formula.Formula,smile.data.DataFrame,DataFrameRegression> trainer) Runs bootstrap tests.
static <T> double[]	regression(int k, T[] x, double[] y, java.util.function.BiFunction<T[],double[],Regression<T>> trainer) Runs bootstrap tests.
<T> double[]	regression(T[] x, double[] y, java.util.function.BiFunction<T[],double[],Regression<T>> trainer) Runs bootstrap tests.

Methods inherited from class java.lang.Object

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

Field Detail

k

public final int k

The number of rounds of cross validation.

train

public final int[][] train

The index of training instances.

test

public final int[][] test

The index of testing instances.

Constructor Detail

Bootstrap

public Bootstrap(int n,
                 int k)

Constructor.

Parameters:

n - the number of samples.

k - the number of rounds of bootstrap.

Method Detail

classification

public <T> double[] classification(T[] x,
                                   int[] y,
                                   java.util.function.BiFunction<T[],int[],Classifier<T>> trainer)

Runs cross validation tests.

Returns:

the error rates of each round.

classification

public double[] classification(smile.data.formula.Formula formula,
                               smile.data.DataFrame data,
                               java.util.function.BiFunction<smile.data.formula.Formula,smile.data.DataFrame,DataFrameClassifier> trainer)

Runs cross validation tests.

Returns:

the error rates of each round.

regression

public <T> double[] regression(T[] x,
                               double[] y,
                               java.util.function.BiFunction<T[],double[],Regression<T>> trainer)

Runs bootstrap tests.

Returns:

the root mean squared error of each round.

regression

public double[] regression(smile.data.formula.Formula formula,
                           smile.data.DataFrame data,
                           java.util.function.BiFunction<smile.data.formula.Formula,smile.data.DataFrame,DataFrameRegression> trainer)

Runs bootstrap tests.

Returns:

the root mean squared error of each round.

classification

public static <T> double[] classification(int k,
                                          T[] x,
                                          int[] y,
                                          java.util.function.BiFunction<T[],int[],Classifier<T>> trainer)

Runs cross validation tests.

Returns:

the error rates of each round.

classification

public static double[] classification(int k,
                                      smile.data.formula.Formula formula,
                                      smile.data.DataFrame data,
                                      java.util.function.BiFunction<smile.data.formula.Formula,smile.data.DataFrame,DataFrameClassifier> trainer)

Runs cross validation tests.

Returns:

the error rates of each round.

regression

public static <T> double[] regression(int k,
                                      T[] x,
                                      double[] y,
                                      java.util.function.BiFunction<T[],double[],Regression<T>> trainer)

Runs bootstrap tests.

Returns:

the root mean squared error of each round.

regression

public static double[] regression(int k,
                                  smile.data.formula.Formula formula,
                                  smile.data.DataFrame data,
                                  java.util.function.BiFunction<smile.data.formula.Formula,smile.data.DataFrame,DataFrameRegression> trainer)

Runs bootstrap tests.

Returns:

the root mean squared error of each round.