package jsat.regression;
import jsat.classifiers.CategoricalData;
import jsat.classifiers.DataPoint;
import jsat.exceptions.FailedToFitException;
/**
* UpdateableRegressor is an interface for one type of Online learner. The main
* characteristic of an online learner is that new example points can be added
* incrementally after the classifier was initially trained, or as part of its
* initial training.
* Some Online learners behave differently in when they are updated. The
* UpdateableRegressor is an online learner that specifically only performs
* additional learning when a new example is provided via the
* {@link #update(jsat.classifiers.DataPoint, double) } method.
*
* The standard behavior for an UpdateableRegressor is that the user first
* calls {@link #train(jsat.regression.RegressionDataSet) } to first train
* the classifier, or {@link #setUp(jsat.classifiers.CategoricalData[], int) }
* to prepare for online updates. Once one
* of these is called, it should then be safe to call
* {@link #update(jsat.classifiers.DataPoint, double) } without getting a
* {@link FailedToFitException}. Some online learners may require one of the
* train methods to be called first.
*
* @author Edward Raff
*/
public interface UpdateableRegressor extends Regressor
{
/**
* Prepares the classifier to begin learning from its
* {@link #update(jsat.classifiers.DataPoint, double) } method.
*
* @param categoricalAttributes an array containing the categorical
* attributes that will be in each data point
* @param numericAttributes the number of numeric attributes that will be in
* each data point
*/
public void setUp(CategoricalData[] categoricalAttributes, int numericAttributes);
/**
* Updates the classifier by giving it a new data point to learn from.
* @param dataPoint the data point to learn
* @param targetValue the target value of the data point
*/
public void update(DataPoint dataPoint, double targetValue);
@Override
public UpdateableRegressor clone();
}
Ads help maintain this website.