Updates in 4 Extensions - Toolbox, SMILE, Converters

MartinLiebig

Over the summer the Data Science teams at RapidMiner were hard at work updating several of our extensions. We are proud to announce updates to the following:

•                Operator Toolbox
•                SMILE
•                Converters
•                Prescriptive Analytics Extension

Here’s a quick overview of what’s new.

Summary Changelog

Operator Toolbox Extension:

·        New operator: GLM Contribution

·        New operator: Generate Partial Dependency Plot Data

·        New operator: Sample Collection

·        Enhancement: Apply Model (Documents) (improved meta data propagation)

Smile Extension:

·        New operator: Compare Distributions

·        New operator: Gradient Boosted Tree (Smile)

·        Major internal code refactoring. Previously trained models may not be applicable anymore.

Converters Extension:

·        New operator: Extract Tokens, which extracts tokens of a document

·        Bugfix in Logistic Regression to ExampleSet, so that the Operator now also can handle Generalized Linear Mode

Prescriptive Analytics Extension:

·        Better and proper error messages in ‘Prescriptive Optimization’

 

Please keep in mind that these extensions are not officially supported. We, as a team, may sometimes make changes which are not backwards compatible! Check out our blog post “5 New Operators Added to the Operator Toolbox and Smile Extensions” for more detailed information.

hughesfleming68

Smile Gradient Boosted Trees vs H2O GBT........Smile GBT is so fast. Well worth checking out.

MartinLiebig

Hey @hughesfleming68,
this is pretty much how i reacted. Also their RF is super quick (ours will get a bug fix in 9.4)
The 'down-side' of SMILEs implementation is usually, that they only have limited parameters. Maybe that is not a down-side though.

Cheers,
Martin