Java* API Reference for Intel® Data Analytics Acceleration Library 2019

List of all members
DistributedStep1Local Class Reference

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

Class Constructor

Constructs the neural network training algorithm by copying input objects and parameters of another algorithm

Parameters
contextContext to manage the algorithm
otherAn algorithm to be used as the source to initialize the input objects and parameters of the algorithm
DistributedStep1Local ( DaalContext  context,
Class<?extends Number >  cls,
TrainingMethod  method 
)

Constructs the neural network training algorithm

Parameters
contextContext to manage the algorithm
clsData type to use in intermediate computations for the algorithm, Double.class or Float.class
methodComputation method of the algorithm, TrainingMethod

Constructs neural network model-based training object with Float data type used for intermediate computations and default computation method

Parameters
contextContext to manage the neural network training object

Detailed Description

Member Function Documentation

DistributedStep1Local clone ( DaalContext  context)

Returns the newly allocated neural network training algorithm with a copy of input objects and parameters of this neural network training algorithm

Parameters
contextContext to manage the algorithm
Returns
The newly allocated algorithm
PartialResult compute ( )

Runs the step 1 of the neural network training algorithm

Returns
Partial results of the neural network training algorithm
TrainingResult finalizeCompute ( )

Computes the results of the neural network training algorithm

Returns
Results of the neural network training algorithm
void setPartialResult ( PartialResult  partialResult)

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

Parameters
partialResultStructure to store partial results of the neural network training algorithm
void setResult ( TrainingResult  result)

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

Parameters
resultStructure to store the results of the neural network training algorithm

Member Data Documentation

Input data

TrainingParameter parameter

Parameters of the algorithm


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

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