Yale Data Clustering Package
 |
The Yale Clustering package provides several data
clustering related algorithms in a pure Java implementation.
Clustering
is a data mining task that allows to analyze data in an explorative
way. It is used in many domains ranging from image processing to
customer segmentation for marketing purposes.
All algorithms and methods
provided by the clustering package are fully integrated into the RapidMiner environment, allowing to combine clustering with many other
preprocessing and data mining methods. Also the package contains
several visualizations for similarities and cluster models, allowing to
configure and inspect the clustering process interactively.
|
|
The key features are
- 100% Java implementation
- Simple API for clustering and similarity
calculation
- Very easy to configure and extend
- Flexible choice of processing steps
- Integrates with all RapidMiner data mining
algorithms
(e.g. evolutionary feature selection)
- Provides various cluster and similarity
visualizations
|
New in Version 1.1:
- K-distance visualization
- Cluster characterization
- Text clustering support
- Library of similarity and distance measures
|
Download
The Yale Clustering Package is now fully integrated into the Rapid Miner
environment.
Screenshots
|
The Clustering package allows to visualize cluster models
in diverse ways, e.g. by a textual description or as trees, as you can
see on the left side.
|
|
Another possible visualization is as graph. Several
layout algorithms are supported. The graph can be explored by zooming
and scaling. The result can be exported as graphics file..
|
|
The Clustering package supports the definition and
analysis of similarity measures. On the left, you see for instance a
histogram of similarities.
|
 |
Clustering algorithms can be combined with all kinds of Yale operators, for instance preprocessing and classification.
|
Contact
The Yale Clustering package is now fully integrated into RapidMiner. For support please refer to the Rapid Miner homepage.
|
|
