Cluster Reinforcement (CR) phase (Scripts) Publisher's description
from Narine Manukyan
The Cluster Reinforcement phase advances cluster separation in a self-organizing map (SOM) by strengthening cluster boundaries in a data-driven manner.
The Cluster Reinforcement phase advances cluster separation in a self-organizing map (SOM) by strengthening cluster boundaries in a data-driven manner. SOM is a self-organized projection of high dimensional data onto a typically two dimensional (2D) feature map, wherein vector similarity is implicitly translated into topological closeness in the 2D projection. The CR phase amplifies within-cluster similarity in an unsupervised, data-driven manner. Discontinuities in the resulting map correspond to between-cluster distances and are stored in a boundary (B) matrix. CR phase enables a new hierarchical visualization of cluster boundaries displayed directly on feature maps, which requires no further clustering beyond what was implicitly accomplished during self-organization in SOM training.
NOTE: CODE FOR B-MATRIX VISUALIZATION AND DEMO CALLS TO CR PHASE WILL BE POSTED SOON. 8 Mar 2012.
PLEASE REFERENCE THE FOLLOWING IF YOU USE THIS CODE:
Manukyan N., Eppstein, M.J., Rizzo, D.M., "Data-Driven Cluster
Reinforcement and Visualization in Sparsely-Matched Self-Organizing
Maps", IEEE Transactions on Neural Networks and Machine Learning, 2012 (in press).
System Requirements:No special requirements.
Program Release Status: New Release
Program Install Support: Install and Uninstall