C++ API Reference for Intel® Data Analytics Acceleration Library 2019 Update 5

List of all members
Distributed< step2Master, algorithmFPType, method > Class Template Reference

Provides methods for neural network model-based training in the batch processing mode. More...

Class Declaration

template<typename algorithmFPType, Method method>
class daal::algorithms::neural_networks::training::interface1::Distributed< step2Master, algorithmFPType, method >

Template Parameters
algorithmFPTypeData type to use in intermediate computations for neural network model-based training, double or float
methodNeural network training method, training::Method
Enumerations
References

Constructor & Destructor Documentation

Distributed ( const Distributed< step2Master, algorithmFPType, method > &  other)
inline

Constructs neural network by copying input objects and parameters of another neural network

Parameters
[in]otherAn algorithm to be used as the source to initialize the input objects and parameters of the algorithm

Member Function Documentation

services::SharedPtr<Distributed<step2Master, algorithmFPType, method> > clone ( ) const
inline

Returns a pointer to the newly allocated neural network with a copy of input objects and parameters of this neural network

Returns
Pointer to the newly allocated layer
virtual int getMethod ( ) const
inlinevirtual

Returns method of the algorithm

Returns
Method of the algorithm
DistributedPartialResultPtr getPartialResult ( )
inline

Returns structure that contains computed partial results of the neural network algorithm

Returns
Structure that contains partial results of the neural network algorithm
ResultPtr getResult ( )
inline

Returns the structure that contains the results of the neural network algorithm

Returns
Structure that contains the results of the neural network algorithm
services::Status initialize ( const services::Collection< size_t > &  dataSize,
const training::Topology &  topology 
)
inline

Initializes the neural network topology

Parameters
[in]dataSizeCollection of sizes of each dimension of input data tensor
[in]topologyNeural network topology
Returns
Status of computations
services::Status setPartialResult ( const DistributedPartialResultPtr &  partialResult)
inline

Registers user-allocated memory to store partial results of the neural network algorithm

Parameters
[in]partialResultStructure for storing partial results of the neural network algorithm
Returns
Status of computations

Member Data Documentation

Input data structure

ParameterType parameter

Training parameters


The documentation for this class was generated from the following file:

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