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

List of all members
TrainingDistributed Class Referenceabstract

Provides methods to train models that depend on the data provided in the distributed processing mode. For example, these methods enable training the linear regression model. Classes that implement specific algorithms of model training in the distributed processing mode are derived classes of the TrainingDistributed class. The class additionally provides methods for validation of input and output parameters of the algorithms. More...

Class Constructor

Constructs the training algorithm in the distributed processing mode

Parameters
contextContext to manage the training algorithm in the distributed processing mode

Detailed Description

Member Function Documentation

void checkComputeParams ( )

Validates parameters of the compute method

void checkFinalizeComputeParams ( )

Validates parameters of the finalizeCompute method

abstract TrainingDistributed clone ( DaalContext  context)
abstract

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

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

Computes partial results of the algorithm in the distributed processing mode

Returns
Partial results of the algorithm
void dispose ( )

Releases memory allocated for the native algorithm object

Implements Disposable.

Result finalizeCompute ( )

Computes final results of the algorithm using partial results in the distributed processing mode.

Returns
Final results 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.