Autoregressive model definition of the time series at presence of hindrances

The questions of the building of difference schemes in the manner of autoregressive models of time series at presence normal mode and multiplicative noises are considered. Normal-mode interference can be eliminated by means of “forming filter”, adjusted on given type of hindrance. It is shown that presence of white noise brings about offset of estimations of factors of difference schemes. The ways of the removal of bias of estimations are considered. It is stated that addition and multiplicative forms of manifestation of white noise are enough close, and these models are possible if required mutually change.