org.apache.flink.streaming.examples.ml

Class IncrementalLearningSkeleton

java.lang.Object

org.apache.flink.streaming.examples.ml.IncrementalLearningSkeleton

public class IncrementalLearningSkeleton
extends Object

Skeleton for incremental machine learning algorithm consisting of a pre-computed model, which gets updated for the new inputs and new input data for which the job provides predictions.

This may serve as a base of a number of algorithms, e.g. updating an incremental Alternating Least Squares model while also providing the predictions.

This example shows how to use:

• Connected streams
• CoFunctions
• Tuple data types

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	IncrementalLearningSkeleton.FiniteNewDataSource Feeds new data for newData.
static class	IncrementalLearningSkeleton.FiniteTrainingDataSource Feeds new training data for the partial model builder.
static class	IncrementalLearningSkeleton.PartialModelBuilder Builds up-to-date partial models on new training data.
static class	IncrementalLearningSkeleton.Predictor Creates newData using the model produced in batch-processing and the up-to-date partial model.

Constructor Summary

Constructors

Constructor and Description
IncrementalLearningSkeleton()

Method Summary

Modifier and Type	Method and Description
static void	main(String[] args)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

IncrementalLearningSkeleton

public IncrementalLearningSkeleton()

Method Detail

main

public static void main(String[] args)
                 throws Exception

Throws:

Exception