|
YALE 3.3 released on August 4th, 2006 |
|
The new version 3.3 of the free data mining and machine learning environment YALE was released on August 4th, 2006. Among other features, this release offers a completely revised visualization engine which is both more efficient and more powerful.
In addition, several new operators are part of the new release:
- Y-AdaBoost
- Y-Bagging
- MultiCriterionDecisionStumps
- RVMLearner
- Gaussian Process Learner
- ExperimentEmbedder
- OperatorEnabler
- ExampleSetJoin
- Numeric2Binary
- Permutation
Probably the most important enhancements were done in the area of data set and model visualization. The new plotters include quartile plots, color histograms, survey plots, Andrews curves, parallel plots, bar plots, box plots, RadViz, GridViz, and SOMs (Self-Organizing Maps). Other visualization related changes are
- (Meta) data views are now backed up by tables which are
much faster than old HTML views
- New (high-dimensional) plotters and jitter function
for plotting added and old plotters revised
- More intelligent availability checks for plotters and
automatic downsampling if number of data points is too
high
- Support for plotting and logging nominal values
and parameters added
- Data set plotters can now also consider feature weights
- New SVM visualizations
Other features include:
- Yale is now also available as Windows executable
- Includes the newest version of Weka
- Search and Replace for XML tab introduced
- All validation operators are now able to optionally produce
the model of the complete data set
- A double click on an operator in tree view now toggles the breakpoint
status
- Users can specify a search string and learner capabilities in the
new operator dialog now; dialog provides an add button now
- New operator tree properties allow filtering of disabled
operators or expansion of the complete tree
- Default file extensions for all IO files now
- Debug mode added
- Support for automatic parameter optimization of
nominal parameters added
- Ensemble and meta learners with more than
one inner operator (base learner) supported now
- Exceptions for feature filters (skip all ... but not ...)
- ...and many more...
Links
|