|
YALE 3.4 Released on October 3rd, 2006 |
|
The new version 3.4 of the free data mining and machine learning environment
YALE was released on October 3rd, 2006.
Among other features, this release offers enhanced input handling methods for various
file formats and database systems.
Together with the powerful visualization engine - which was already completely revised
for YALE 3.3 - and the huge amount of learning and preprocessing methods, YALE 3.4 is
the first choice for the development of real-world data mining applications.
|
Several new operators are part of the new release:
- SPSSExampleSource
- ExcelExampleSource
- MultivariateSeries2WindowExamples
- EvolutionaryParameterOptimization
- Y-NaiveBayes
- IOObjectReader
- IOObjectWriter
- AGA
- YAGGA2
- LiftChart
- ROCChart
- MacroDefinition (see below)
|
|
|
Probably the most important enhancements were done in the area of data set loading. The new input formats include operators for Microsoft Excel and SPSS. Several database systems are now natively supported and database access is much easier than before.
|
Other (data) in- and output related changes are
- ExampleSource now supports zipped data files
- New data representations backed up by non-double arrays requiring less memory
- SimpleExampleSource now automatically guesses all value types
- New Attribute Editor features easing the input format definition:
- Use arbitrary rows as attribute names (useful, for example, for csv files)
- Re-guess all value types
- Data and meta data saving reminder
- Easy definition of JDBC drivers: by specifying them in CLASSPATH
or by copying them into the directory lib/jdbc
- Free JDBC drivers for MySQL, PostgreSQL, Microsoft SQL Server, and Sybase included
- Improved the directly working on database mode (DatabaseES)
- New configuration wizards for ExampleSource and DatabaseExampleSource - drastically easing the usage of these complex and powerful operators
Other features include:
- Added macro definitions: macros can be defined with the operator MacroDefinition and used with %{my_macro}.
- Improved logging
- Operator configuration wizards
- The minimum and maximum colors for plotters can now be specified
- Density and SOM plotters now support example visualization and are more more efficient
- SOM plots are now interpolated
- New operator dialog shows deprecation information
- Allowing both, attribute and example representations for Series 2 Window Examples
- Plotters more robust against missing values
- Improved data saving for ExampleSets
- Includes the newest version of Weka
- A large number of bugfixes
- ...and a lot more...
|
Links
|
|
