Rapid-I released a new version of the leading open-source data mining, ETL, and predictive BI solution RapidMiner 4.4

We are happy to announce the release of RapidMiner 4.4. Since the release 4.3 in November 2008, we again worked hard on improving RapidMiner to meet your data analysis, ETL, and BI requirements. The result is now available for download and provides 35 new operators, more than 30 bugfixes, and a lot of exciting new features:

• New Preprocessing Options: removal of duplicates, nominal value splits, data set union and superset and others
• New Learning Schemes: linear and quadratic discriminant analysis, fast large margin and others
• New Process Possibilities: FileIterator, MacroConstruction, SetData, ExceptionHandling and others
• New Improvements: better handling of time zones, helper operators for attribute renaming and others

You will find detailed information about this release below. Or you can directly proceed to our download section.

More Features and Improved Usability: Changes in the process setup are now allowed even during process runtime, e.g. when waiting at a breakpoint Automatical adaptation of parameters at operator renaming Default parameter values are also stored in the process files now Deprecated operators are now rendered with a gray color New range statistics for nominal attributes New context menu for tables (sorting, column widths...) New Operator dialog now also supports full text search Improved process structure checks Support for different time zones Double click between two columns in table headers resizes the left column, pressing CTRL resized all columns More than 30 new operators!
Improved Visualizations: Graphs now display the strengths of the edges by color Tree nodes do no longer overlap Decision trees now show the subtree size as tool tip for the inner nodes Decision tree edges are now darker for larger subtrees Date and times are now displayed in the system's local settings New plotter: Block Added support for applying a log scale for the color column The bar chart now displays the category as label in the domain axis Added the ability to "explode" tiles of pie and ring charts
Better Modeling: Decision trees learn faster now Series windowing operators now also support nominal attributes NaiveBayes can now handle new nominal values during model application New binominal classification performance measures (positive predictive value, negative predictive value, psep) Improved neural network learning ExampleSetMerge, GuessValueTypes and several other operators can now handle missing values
More Power: MacroConstruction allows for almost arbitrary calculations based on macros ExceptionHandling allows the handling of expected errors and can be especially useful in combination with the IO storage mechanism and process branching ClearProcessLog now also allows to remove the complete logging table GuessValueTypes now only works on regular attributes and provides a parameter for extending it on the special attributes AttributeFilter now also provides a new parameter work_on_special The operator Replace now also allows empty replace_by values All feature selection and weighting operators now allow name logging The Replace operator now supports capturing groups Improved quote handling for file based data loading
Less Bugs: Fixed GUI related memory leaks Fixed bug in meta data table viewer in cases where huge numbers of long nominal values existed which caused a crash of the Java Virtual Machine in some cases Fixed bug in Series Multiple Plotter which prevented rescaling Special format in ExampleSetWriter did not quote whitespaces Fixed several DBScan issues Simple sampling did not obey local random seeds Process Logging to files prevented the writing of the first logged result Binary performance measure was not delivering the fitness for specificity, sensitivity,and youden index Fixed bug in parameter list of ValueSubgroupIterator ViewAttributes did not add min and max statistics so that those statistics where not calculated on data table views Fixed bug in Windows GUI start script (linebreak) Fixed bug for surface 3D plot where x and y were replaced by each other Fixed potential thread deadlock during the filling of the plotter list Fixed a bug in ChiSquaredWeighting for mixed-type data sets where the number of bins was smaller than the maximum number of nominal values The default global random seed in the preferences dialog was not allowed to be set to -1 The property keys of the preferences dialog were editable Fixed bug in PolynomialRegression And several more...

RapidMiner 4.4: Enterprise Edition

The new version of course is again available in different versions. Users can choose between the free Community Edition of RapidMiner and one of the Enterprise Editions which are more suitable for professional data analysts:

      Enterprise Edition = 
            Community Edition +
            More Features +
            Services +
            Guarantees

Of course, the RapidMiner Enterprise Edition benefits from all improvements stated above. In addition, the reporting engine was again extended and further improved and for example now allow a better corporate design, a default logo and more reporting settings.

Smaller companies and inviduals can use the new RapidMiner Small Enterprise Edition which contains all additional functionality from the RapidMiner Enterprise Edition including the reporting engine and the support of multi-core systems which leads to drastic increases in analysis speed. The RapidMiner Standard Enterprise Edition additionally provides extended guarantees and services including the support for optimizing your analysis processes. Companies can get the full service level together with the extended features for an unlimited number of users with the RapidMiner Unlimited Enterprise Edition - for a price less than a single program license of other proprietary solutions. Check out the details at the Enterprise Edition feature page!

Links