We consider the regression model

\[y = X \beta + \mu\] \[\mu \sim N(0, \sigma^2 \Omega(\theta))\]

So, we have to estimate $\beta, \sigma^2, \theta$

We write \(E(\mu(t)|\mu)=H(t)=cov(\mu(t), \mu)var(\mu)^{-1}\)

\[\hat y(t, \hat \theta) = X \hat \beta(\hat \theta) +\]

---

Bibliography

Kohn R. and Ansley C. (1985). Efficient estimation and prediction in time series regression models. Biometrika (1985), 72, 3, pages 694-7.