Authors: Tang Van To and S. Thummaroj
The Kohonen Network is a powerful tool in exploratory phase of data mining. It is capable of projecting high dimensional data onto a low, usually 2-dimensional grid of neurons with good neighborhood preservation and a good tool for dimensional reduction. However, when to due with the large quantities of high-dimensional data, a large size of networks should be used and it need an extremely time consuming. Hierarchical Kohonen Network is a combination of multi branching decision tree and Kohonen Network, one of solution to overcome the drawback of Kohonen Network. A comparison between Kohonen Network and Hierarchical Kohonen Network is done and a possible way to enlarge the size of Kohonen Network through Hierarchical Kohonen Network will be discussed.