org.deeplearning4j.earlystopping.scorecalc

Interface ScoreCalculator<T extends Model>

Type Parameters:

T - Type of model. For example, MultiLayerNetwork or ComputationGraph

All Superinterfaces:

Serializable

All Known Implementing Classes:

AutoencoderScoreCalculator, BaseIEvaluationScoreCalculator, BaseMLNScoreCalculator, BaseScoreCalculator, ClassificationScoreCalculator, DataSetLossCalculator, DataSetLossCalculatorCG, RegressionScoreCalculator, ROCScoreCalculator, VAEReconErrorScoreCalculator, VAEReconProbScoreCalculator

public interface ScoreCalculator<T extends Model>
extends Serializable

ScoreCalculator interface is used to calculate a score for a neural network. For example, the loss function, test set accuracy, F1, or some other (possibly custom) metric.

Method Summary

Modifier and Type	Method and Description
double	calculateScore(T network) Calculate the score for the given MultiLayerNetwork
boolean	minimizeScore()

Method Detail

calculateScore

double calculateScore(T network)

Calculate the score for the given MultiLayerNetwork

minimizeScore

boolean minimizeScore()

Returns:

If true: the score should be minimized. If false: the score should be maximized.