Module org.nasdanika.ai

Package org.nasdanika.ai

Class AbstractRecursiveDoubleFitter

java.lang.Object

org.nasdanika.ai.AbstractDoubleFitter

org.nasdanika.ai.AbstractRecursiveDoubleFitter

All Implemented Interfaces:

FittedPredictor.Fitter<double[],double[],Double>

public abstract class AbstractRecursiveDoubleFitter
extends AbstractDoubleFitter

Creates a Function<double[][],double[]> predictor for each label element with features for later elements including labels for earlier. During prediction earlier predictors outputs are used as inputs for later predictors. This is essentially how autoregression works.

Constructor Summary

Constructors

Constructor	Description
AbstractRecursiveDoubleFitter()

Method Summary

Modifier and Type	Method	Description
protected abstract Function<double[][],double[]>	fit(double[][] features, double[] labels)
protected final Function<double[][],double[][]>	fit(double[][] features, double[][] labels)

Methods inherited from class org.nasdanika.ai.AbstractDoubleFitter

compose, computeError, createPredictor, fit, fitAsync, wrap

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.nasdanika.ai.FittedPredictor.Fitter

adaptFeature, adaptFeatureAsync, adaptLabel, adaptLabelAsync, compose

Constructor Details

AbstractRecursiveDoubleFitter

public AbstractRecursiveDoubleFitter()

Method Details

fit

protected final Function<double[][],double[][]> fit(double[][] features,
 double[][] labels)

Specified by:

fit in class AbstractDoubleFitter

fit

protected abstract Function<double[][],double[]> fit(double[][] features,
 double[] labels)