Development and research of nonparametric estimationof probability density based on the principle of training sample decomposition according to its volume

The two-level nonparametric estimation of probability density based on the principle of decomposition of training sample according to its volume is offered. This estimation is characterized by high computing efficiency due to the opportunity of parallel computing technologies use. The results of computing experiments are given.

estimation of density probability, the big samples, asymptotic properties, research of the properties, parallel computing technologies