Visual Clusters is the core module of the Viscovery SOMine suite. It can be operated as a stand-alone system, as well as be flexibly extended with other modules of the suite.
The two basic Preprocess and Cluster workflows are available; the first guiding the user through data import and preprocessing, the latter through the creation and exploration of the SOM model.
Data sets with up to 100,000 records and up to 100 variables can be processed.
In addition to the capabilities listed in the Viscovery SOMine page, Visual Clusters provides all basic tools for SOM creation and visualization, interactive exploration, and visual cluster analysis.
Main functions and features
- Multiple data preprocessing capabilities
- Definition and automatic management of nominal variables
- Tools for adjustment of histograms
- Creation of self-organizing map models using predefined schedules
- Interactive SOM visualization and exploration
- Visual cluster analysis with integrated visualization of cluster boundaries and inner structures
- Display of thumbnails from external documents over the map window
- Statistical functions, such as descriptive statistics, histograms, correlations, PCA, and scatter plots
- Cross-references to original data