jsat.regression

Interface WarmRegressor

All Superinterfaces:

java.lang.Cloneable, Regressor, java.io.Serializable

All Known Implementing Classes:

DCDs, LinearBatch, LSSVM, PlattSMO, SDCA

public interface WarmRegressor
extends Regressor

This interface is meant for models that support efficient warm starting from the solution of a previous model. Training with a warm start means that instead of solving the problem from scratch, the code can use a previous solution to start closer towards its goal.

Some algorithm may be able to warm start from solutions of the same form, even if they were trained by a different algorithm. Other algorithms may only be able to warm start from the same algorithm. There may also be restrictions that the warm start can only be from a solution trained on the exact same data set. The latter case is indicated by the warmFromSameDataOnly() method.

Just because a regressor fits the type that the warm start interface states doesn't mean that it is a valid classifier to warm start from. Regressors of the same class trained on the same data must always be valid to warm start from.

Note: The use of this class is still under development, and may change in the future.

Method Summary

Modifier and Type	Method and Description
void	train(RegressionDataSet dataSet, Regressor warmSolution) Trains the regressor and constructs a model for regression using the given data set.
void	train(RegressionDataSet dataSet, Regressor warmSolution, boolean parallel) Trains the regressor and constructs a model for regression using the given data set.
boolean	warmFromSameDataOnly() Some models can only be warm started from a solution trained on the exact same data set as the model it is warm starting from.

Methods inherited from interface jsat.regression.Regressor

clone, regress, supportsWeightedData, train, train

Method Detail

warmFromSameDataOnly

boolean warmFromSameDataOnly()

Some models can only be warm started from a solution trained on the exact same data set as the model it is warm starting from. If this is the case true will be returned. The behavior for training on a different data set when this is defined is undefined. It may cause an error, or it may cause the algorithm to take longer or reach a worse solution.
When true, it is important that the data set be unaltered - this includes mutating the values stored or re-arranging the data points within the data set.

Returns:

true if the algorithm can only be warm started from the model trained on the exact same data set.

train

void train(RegressionDataSet dataSet,
           Regressor warmSolution,
           boolean parallel)

Trains the regressor and constructs a model for regression using the given data set. If the training method knows how, it will used the threadPool to conduct training in parallel. This method will block until the training has completed.

Parameters:

dataSet - the data set to train on

warmSolution - the solution to use to warm start this model

parallel - true if the training should be done using multiple-cores, false for single threaded.

train

void train(RegressionDataSet dataSet,
           Regressor warmSolution)

Trains the regressor and constructs a model for regression using the given data set.

Parameters:

dataSet - the data set to train on

warmSolution - the solution to use to warm start this model