Interface and Description |
---|
org.nd4j.linalg.learning.GradientUpdaterAggregator
Use Updater view array functionality instead
|
Class and Description |
---|
org.nd4j.linalg.lossfunctions.LossCalculation
Use
ILossFunction for equivalent functionality.
Allows for custom loss functions. Use computeScore with an instance of the loss function classs instead. |
org.nd4j.linalg.dataset.api.iterator.MultipleEpochsIterator |
org.nd4j.linalg.dataset.api.iterator.StandardScaler
Use
NormalizerStandardize for equivalent functionality.
NormalizerStandardize is more stable (for examples, when a column contains all the same values for every example) but
otherwise provides equivalent functionality. See also NormalizerMinMaxScaler |
Enum Constant and Description |
---|
org.nd4j.linalg.lossfunctions.LossFunctions.LossFunction.CUSTOM |
org.nd4j.linalg.lossfunctions.LossFunctions.LossFunction.EXPLL |
org.nd4j.linalg.lossfunctions.LossFunctions.LossFunction.RMSE_XENT |
Copyright © 2017. All Rights Reserved.