Implementation of the Mountain Clustering Method and Comments on its Practical use for Determining Cluster Centers
MetadataShow full item record
For certain applications a need arises to reduce a large set of measurement data and select a group of the most representative data. Such a situation occurs for example in the case of the fuzzy logic algorithms whose computational complexity makes them inapplicable to too large input sets. One of the methods to reduce a data set is to determine the centre of clusters, that is the elements being the optimum representation of the entire set. The purpose of this paper is to describe the operation of the potential method designed to locate the centres of clusters in the entire set of measurement data. We present the selection algorithm based on the assumption that in certain local environments data are normally distributed. This assumption proves to be correct in numerous practical applications; however, in some cases a different probability distribution may seem more appropriate. For these cases we will only hint at how one can try to modify the potential function to produce the most reliable effect. Along with the mathematical description of the method we also present the functionality of a dedicated software implemented for this purpose.
- Artykuły / Articles 
Using this material is possible in accordance with the relevant provisions of fair use or other exceptions provided by law. Other use requires the consent of the holder.