The problem of merging Gaussian mixture components is discussed in situations where a Gaussian mixture is fitted but the mixture components are not separated enough from each other to interpret them as "clusters". Two methods are introduced, corresponding to two different "cluster concepts" (separation by gaps and "data patterns"). A visualisation of the modality of a density of a mixture of two Gaussians is proposed and the stability of the unmerged Gaussian mixture is compared to that of clusterings obtained by merging components. © Springer-Verlag Berlin Heidelberg 2010.
CITATION STYLE
Hennig, C. (2010). Ridgeline plot and clusterwise stability as tools for merging Gaussian mixture components. In Studies in Classification, Data Analysis, and Knowledge Organization (pp. 109–116). Kluwer Academic Publishers. https://doi.org/10.1007/978-3-642-10745-0_11
Mendeley helps you to discover research relevant for your work.