High dimensional datasets, such as text, are very often hard to cluster, and autonomous clustering of the dataset is even more taxing. Many algorithms have been developed and utilized, such as neural networks and classifier systems, endeavoring to cluster these datasets; however, most implementations require domain knowledge of the collection. A viable, more complete alternative to these systems is the self-organizing map.7.5.3 Clustering Method The final method of determining the BMU and neighborhood is a hybridized version of the two ... the membership of the clusters , and the neighborhood radius of the connection is enlarged if needed; this ensures furtheranbsp;...
|Title||:||Solving the Segmented, Static Database Paradigm by Means of Prismal Self-organizing Maps|
|Publisher||:||ProQuest - 2007|