scientific article; zbMATH DE number 3874449
Let \(X(s)=(X_ 1(s),...,X_ n(s))\) and \(Y(s)=(Y_ 1(s),...,Y_ n(s))\) be two n-dimensional stationary processes such that \(Y(s)=\sum^{N}_{\nu =0}a_{\nu}X(s-\nu\theta )\), where \(a_ 0=1\), \(\theta >0\) and \(z^ N+a_ 1z^{N-1}+...+a_ N\neq 0\) for \(| z|\geq 1.\) Under some general conditions the author derives procedures for finding (i) the best linear estimator of \(Y_ 1(t+\tau)\) given \(\{\) Y(s),t-\(T\leq s\leq t\}\), and (ii) the best linear estimator of \(X_ 1(t+\tau)\) given \(\{\) X(s),t-\(T\leq s\leq t\}\). The procedures are based on the Yaglom's method for the extrapolation of continuous multidimensional stationary processes.
- Following a trend with an exponential moving average: analytical results for a Gaussian model
- scientific article; zbMATH DE number 3846577 (Why is no real title available?)
- Forecast Functions Implied by Autoregressive Integrated Moving Average Models and Other Related Forecast Procedures
- scientific article; zbMATH DE number 4058701 (Why is no real title available?)
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3339979)