Clustering With Prototype Entity Selection Compared With K-Means

Abstract

Clustering is an important area of application for a variety of fields including data mining. This paper presents a new clustering method namely, the Clustering with Prototype Entity Selection (abbreviated CPES), proposed as a method of clustering for data mining. The CPES method is original to the authors. The paper describes its mathematical essence, presents the algorithm and the experimental results obtained as compared to the K-Means method. The K-Means algorithm is by far the most widely used method for discovering clusters in data.