Using the fact that a multivariate random sample of n observations also generates\nn nearest neighbour distance (NND) univariate observations and\nfrom these NND observations, a set of n auxiliary observations can be obtained\nand with these auxiliary observations when combined with the original\nmultivariate observations of the random sample, a class of pseudodistance\nh D is allowed to be used and inference methods can be developed using this\nclass of pseudodistances. The h D estimators obtained from this class can\nachieve high efficiencies and have robustness properties. Model testing also\ncan be handled in a unified way by means of goodness-of-fit tests statistics\nderived from this class which have an asymptotic normal distribution. These\nproperties make the developed inference methods relatively simple to implement\nand appear to be suitable for analyzing multivariate data which are often\nencountered in applications.
Loading....