ABSTRACT
Many of them have been used in data mining and several are machine learning approaches. Support vector clustering Support vectoring clustering (SVC) algorithms have arisen from support vector machines which are supervised learning models with learning algorithms used to analyse data used for classification. SVC is carried out in two stages. The first stage involves describing the clusters, namely, where the minimal enclosing hypersphere of the data is computed by finding the support vectors which define the cluster boundaries and it uses a similar constraint optimisation approach as that of support vector machines. During the second stage, clusters are labelled such that patterns are assigned to clusters. The two stages are carried out iteratively in order to tune the hyper-parameters of SVC. In the literature different Self-Organzing Maps (SOMs) for classifying time series have been proposed. The SOMs produce a mapping of high-dimensional input data onto units of a low-dimensional lattice that is ordered and descriptive of distribution of input.
