Answers are grounded in this paper and cite their sources.
Let ₜ\ be a linear stationary process of the form xₜ + Σ1 i<∞aᵢxₜ₋ᵢ = eₜ, where ₜ\ is a sequence of i.i.d. normal random variables with mean 0 and variance σ². Given observations x₁, ⋯, xₙ, least squares estimates â(k) of a' = (a₁, a₂, ⋯), and σ̂²ₖ of σ² are obtained if the kth order autoregressive model is assumed. By using â(k), we can also estimate coefficients of the best predictor based on k successive realizations. An asymptotic lower bound is obtained for the mean squared error of the estimated predictor when k is selected from the data. If k is selected so as to minimize Sₙ(k) = (n + 2k)σ̂²ₖ, then the bound is attained in the limit. The key assumption is that the order of the autoregression of ₜ\ is infinite.
No takes yet. Share an insight, caveat, or question.
Ritei Shibata (1980) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: