Package com.datumbox.framework.core.machinelearning.common.interfaces

Interface Summary

Interface	Description
Cluster	The Cluster object stores all the information of the cluster, including the assigned points, the parameters and the label.
ModelParameters	The ModelParameter objects stores the coefficients that were learned during the training of the algorithm.
Parallelizable	All algorithms capable of using parallelism implement this interface.
PredictParallelizable	All Machine Learning models capable of predicting records in parallel implement this interface.
StepwiseCompatible	Stepwise regression requires the estimation of p-values for each parameter.
Trainable<MP extends Learnable,TP extends Parameterizable>	This interface is used to mark classes that can be trained.
TrainingParameters	The TrainingParameters objects store all the initial parameters provided to the algorithms during training.
TrainParallelizable	All Machine Learning models capable of using parallelism during training implement this interface.
ValidationMetrics	Interface for every Metrics class in the framework.

Class Summary

Class	Description
PredictParallelizable.Prediction	Tuple that stores the results of the prediction namely the yPredicted and the yPredictedProbabilities objects.