C++ API Reference for Intel® Data Analytics Acceleration Library 2020 Update 1

Namespaces | Enumerations
daal::algorithms::neural_networks::layers::backward Namespace Reference

Contains classes for the backward stage of the neural network layer.

Namespaces

 interface1
 Contains version 1.0 of Intel(R) Data Analytics Acceleration Library (Intel(R) DAAL) interface.
 

Enumerations

enum  InputId { inputGradient }
 
enum  InputLayerDataId { inputFromForward = lastInputId + 1 }
 
enum  ResultId { gradient, weightDerivatives, biasDerivatives }
 
enum  ResultLayerDataId
 

Enumeration Type Documentation

enum InputId

Available identifiers of input objects for the layer algorithm

Enumerator
inputGradient 

Gradient of the preceding layer

Available identifiers of input objects for the layer algorithm

Enumerator
inputFromForward 

Input structure retrieved from the result of the forward layer

enum ResultId

Available identifiers of results for the layer algorithm

Enumerator
gradient 

Gradient with respect to the outputs

weightDerivatives 

Gradient with respect to the weight

biasDerivatives 

Gradient with respect to the bias

Available identifiers of result objects for the layer algorithm

For more complete information about compiler optimizations, see our Optimization Notice.