A novel classifier combining supervised and unsupervised learning methods

Nearest Centroid Neighbor (NCN) classifier is a fast and simple algorithm representing supervised methods of data classification. This algorithm assumes that all classes can be represented by the individual clusters and the classes means (centroids) are used to determine to which class a new unknown sample belongs. However the assumption that each class consists of one and exactly one cluster limits the performance of the algorithm. A method of replacing this one cluster by two or more subclusters is proposed in this paper. The new algorithm uses a well-known unsupervised classification method k-means clustering to find the centroids of these subclusters. This modification of NCN classifier improves the results achieved on several databases