Abstract | ||
---|---|---|
Self-Organizing Maps (SOM) are very powerful tools for data mining, in p articular for visualizing the distribution of the data in v ery high- dimensional data sets. Moreover, the 2D map produced by SOM can be used for unsupervised partitioning of the original data set i nto categories, provided that this map is s omehow adequately segmented in clusters. This is usually done either manually by visual inspection, or by applying a classical clustering technique (such as agglomerative clustering) to the set of prototypes corresponding to the map. In this paper, we present a new approach for the segmentation of Self- Organizing Maps after training, which is both very simple a nd efficient. Our algorithm is based on a post-processing of the U-matrix (the matrix of distances between adjacent map units), which is directly derived from an elementary image-processing technique. It i s s hown on some simulated data sets that our partitioning algorithm appears to give very good results in terms of segmentation quality. Preliminary results on a real data set also seem to indicate that our algorithm can produce meaningful clusters on real data. |
Year | Venue | Keywords |
---|---|---|
2004 | ESANN | high dimensional data,visual inspection,data mining |
Field | DocType | Citations |
Cluster (physics),Data mining,Data set,Visual inspection,Matrix (mathematics),Computer science,Self-organizing map,Artificial intelligence,Cluster analysis,Hierarchical clustering,Pattern recognition,Segmentation,Algorithm,Machine learning | Conference | 7 |
PageRank | References | Authors |
0.64 | 6 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
David Opolon | 1 | 8 | 1.02 |
Fabien Moutarde | 2 | 54 | 15.26 |