The paper proposes a new method to control the level of separation of components using a single parameter. An illustration for the latent class model (mixture of conditionally independent multinomial distributions) is provided. Further extensions to other finite mixture models are discussed.
CITATION STYLE
Dias, J. G. (2004). Controlling the Level of Separation of Components in Monte Carlo Studies of Latent Class Models. In Classification, Clustering, and Data Mining Applications (pp. 77–84). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-17103-1_9
Mendeley helps you to discover research relevant for your work.