The exponential statistical experiments with persistent linear regression and equilibrium

摘要

The asymptotic behavior of exponential statistical experiments set by the exponential symmetric statistics is studied. In this case, the property of a persistent linear regression, which means the dependence of separate results of the experiment on the averaging of previous experiments, is assumed. For the exponential statistical experiments in a scheme of series (as N → ∞), the equilibrium mode is established, and the approximation of the normal autoregression by an exponential process is constructed.