ANALYSIS OF PROPERTIES OF NONPARAMETRIC ESTIMATES OF THE PROBABILITY DENSITIES MIX UNDER DIFFERENT CONDITIONS OF STATISTICS DATA DISTRIBUTION
The article considers nonparametric estimation of probability densities mix of Rosenblatt – Parzen type, which synthesis is based on the decomposition of the given statistical data on the volume. The dependence of their properties on the number of the mix components and the level of uneven distribution of observations has been investigated.
Keywords: probability density, nonparametric estimation, decomposition of statistic data, asymptotic properties.