Advertisement
Not a member of Pastebin yet?
Sign Up,
it unlocks many cool features!
- \begin{problem}{1}
- Consider a process of the form
- $$Y_t = R \cos [2\pi(ft + \Phi)], \quad t = 0; \pm1;\pm 2 \dots, $$
- where $0<f< 1/2$ is deterministic, $R$ and $\Phi$ are uncorrelated random variables with $\Phi$ being uniformly distributed in [0, 1].
- \begin{enumerate}[label=\alph*)]
- \item Compute expectation, autocovariance function and autocorrelation function of ($Y_t$)
- \item Check whether the process $Y$ is stationary (in a weak sense).
- \end{enumerate}
- \end{problem}
- \begin{problem}{2}
- Let
- $$X_t = \varphi X_{t-3} + \varepsilon_t,$$
- where ($\varepsilon_t$) is a white noise with $\varepsilon_t ~ N(0;\sigma^2)$.
- \begin{enumerate}[label=\alph*)]
- \item For which values of $\varphi$ is the process ($X_t$) weakly stationary?
- \item Find the autocorrelation function and the spectral density of ($X_t$).
- \end{enumerate}
- \end{problem}
- \begin{problem}{3}
- It is known that a time series $(X_t)_{t \in \mathbb{Z}}$ is stationary with the spectral density of the form
- $$ S(\lambda) = \frac{1}{\pi}\cos^2 (\lambda)(1+cos(\lambda)), -\pi<\lambda<\pi.$$
- Find the autocorrelation function of ($X_t$).
- \end{problem}
- \begin{problem}{4}
- Define a process $(Y_t)_{t \in \mathbb{Z}}$ via
- $$Y_t = X_t + 0.4 X_{t-1} + 0.04 X_{t-2},$$
- where ($X_t$) is a stationary AR(1) process with the autocorrelation function $C_X (h) = 0,5^{|h|}$.
- \begin{enumerate}[label=\alph*)]
- \item Is $(Y_t)_{t \in \mathbb{Z}}$ (weakly) stationary ?
- \item Compute the autocorrelation function of ($Y_t$).
- \item Compute the spectral density of ($Y_t$).
- \end{enumerate}
- \end{problem}
- \begin{problem}{5}
- Let a function $\ro: \mathbb{Z} \implies \mathbb{R}$ is given by
- $$\ro(h) = 1.5(h=0) + 0.75(|h| = 1) + 0.5(|h| = 2).$$
- Is $\ro$ an autocovariance function of some stationary process?
- \end{problem}
- \begin{problem}{6}
- Consider ARMA(1,1) process of the form
- $$X_t = 0.4X_{t-1} + \varepsilon_t + 0.25\varepsilon_{t-1},$$
- where $\varepsilon_t ~ N(0;\sigma^2)$ is a white noise.
- \begin{enumerate}[label=\alph*)]
- \item Check whether the process ($Y_t$) is stationary and invertible.
- \item Compute the spectral density of ($X_t$).
- \end{enumerate}
- \end{problem}
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement