Multivariate statistical experiments with persistent non-linear regression and equilibrium
Dmitri Koroliouk · Theory of Probability and Mathematical Statistics · 2016
A sequence of multivariate statistical experiments with persistent non-linear regression is considered. This sequence is determined by a matrix of directing actions for frequencies of certain attributes. Conditions for the convergence of multivariate statistical experiments to a state of equilibrium are studied. A stochastic approximation by an autoregressive process with normal perturbations is constructed for the discrete time model.