Nonlinear clustering: Methods and applications

Abstract

As a fundamental classification method for pattern recognition, data clustering plays an important role in various fields such as computer science, medical science, social science, and economics. According to the data distribution of clusters, data clustering problem can be categorized into linearly separable clustering and nonlinearly separable clustering. Due to the complex manifold of the real-world data, nonlinearly separable clustering is one of most popular and widely studied clustering problems. This chapter reviews nonlinear clustering algorithms from four viewpoints, namely kernel-based clustering, multi-exemplar model, graph-based method, and support vector clustering (SVC). Accordingly, this chapter reviews four nonlinear clustering methods, namely conscience on-line learning (COLL) for kernel-based clustering, multi-exemplar affinity propagation (MEAP), graph-based multi-prototype competitive learning (GMPCL), and position regularized support vector clustering (PSVC), and demonstrates their applications in computer vision such as digital image clustering, video segmentation, and color image segmentation.