Package org.apache.ignite.ml.environment.parallelism

Package contains realisations of parallelism strategies for multi-thread algorithms.

See: Description

Interface Summary

Interface	Description
ParallelismStrategy	Specifies the behaviour of processes in ML-algorithms that can may be parallelized such as parallel learning in bagging, learning submodels for One-vs-All model, Cross-Validation etc.
Promise<T>	Future interface extension for lambda-friendly interface.

Class Summary

Class	Description
DefaultParallelismStrategy	All task should be processed by default thread pool.
DefaultParallelismStrategy.FutureWrapper<T>	Wrapper for future class.
NoParallelismStrategy	All tasks should be processed in one thread.
NoParallelismStrategy.Stub<T>	Stub for Future interface implementation.

Enum Summary

Enum	Description
ParallelismStrategy.Type	The type of parallelism.

Package org.apache.ignite.ml.environment.parallelism Description

Package contains realisations of parallelism strategies for multi-thread algorithms.