Prescriptive AnalyticsPrescriptive Analytics

This extension offers an operator to do prescriptive optimization. This means you vary the values of an example to optimize a custom fitness function which may derive from a model. Currently supported optimizers: - Grid - Evolutionary - BYOBA

Version 0.16

============

- Added proper error messages if the start value is out of bounds

- The error messages of the inner processes are now properly shown in the GUI.

 

Version 0.1.5

=============

- Added an option to add get candidate results. I.e. more than just the best solution

- Evolutionary is now correctly using custom random seeds - Added Powell as an unbounded optimization method

- A lot of code improvement for optimizer creation and parameter handling

 

Version 0.1.4

=============

- Added "random", "input example" and "reference example" as a starting point options.

- Removed the start setting in SimpleBounds. Please provide start values via input/reference exa. E.g. via Set Data.

- Added a few better error messages for missing references, missing bounds etc

 

Version 0.1.3

=============

- Added a number of interpolation point default boolean for BYOBA

- Added CMA-ES as a optimization method


Product Details

Version 0.1.6
File size 773 kB
Downloads 9070 (20 Today)9070 downloads
Vendor RapidMiner Labs
Category Machine Learning
Released 9/10/19
Last Update 9/10/19 8:00 AM
(Changes)
License AGPL
Product web site rapidminer.com
Rating 0.0 stars(0)