Creating ML Pipelines on Spark

To create a Machine Learning Pipeline for Spark, the developer does the following:

• Divides the pipeline into a logical sequence of components

• Creates or modifies the solution using the Control Center

• Pulls the generated code for the newly added components to his development machine from the code repository

• The generated code will have pre-generated classes for each component, as well as stubs for some methods. The developer provides code for these methods (see here for details)

• Builds and deploys the solution using the Control Center