ALL libraries (COBIB.SI union bibliographic/catalogue database)
  • The REBMIX algorithm for the multivariate finite mixture estimation
    Nagode, Marko ; Fajdiga, Matija
    The article extends the REBMIX to multivariate data. Random variables may follow normal, lognormal, or Weibul parametric families and should be independent within components. The initial weights and ... component parameters are not required. Preprocessing of observations folows the histogram, Parzen window, or k-nearest neighbour approach. The number of components, weights, and component parameters are gained interatively by using information measuresof the distance, such as the total of positive relative deviations andthe information criterion. The number of classes or the number of the nearest neighbours can be optimized, as well. The REBMIX software is availableon http://www.fs.uni-lj.si/lavek
    Source: Communications in statistics. theory and methods. - ISSN 0361-0926 (Vol. 40, iss. 11, 2011, str. 2022-2034)
    Type of material - article, component part
    Publish date - 2011
    Language - english
    COBISS.SI-ID - 11780635
    DOI