Optimal Data Projection for Kernel Spectral Clustering
Diego Hernán Peluffo-Ordóñez, Carlos M. Alzate, Johan A. K. Suykens, German Castellanos-Dominguez · The European Symposium on Artificial Neural Networks · 2014
Spectral clustering has taken an important place in the context of pattern recognition, being a good alternative to solve problems with non-linearly separable groups. Because of its unsupervised nature, clustering methods are often paramet- ric, requiring then some initial parameters. Thus, clustering performance is greatly dependent on the selection of those initial parameters. Furthermore, tuning such parameters is not an easy task when the initial data representation is not adequate. Here, we propose a new projection for input data to improve the cluster identifi- cation within a kernel spectral clustering framework. The proposed projection is done from a feature extraction formulation, in which a generalized distance involv- ing the kernel matrix is used. Data projection shows to be useful for improving the performance of kernel spectral clustering.