1
$\begingroup$

I am interested if following model is stationary,model is represented by following formula

$$ x(n) = \sum_{p=1}^{P} a_p \cos(2\pi f_pn + \phi_p) + \epsilon(n) $$

$n$ is changing from $1$ to $N$, I am reading following material

http://www.stat.duke.edu/courses/Fall99/sta290/Notes/AR/ar.pdf

Main concentration is to understand how good choice is to represent series given by deterministic component plus white noise by AR model, as I know AR model refers to stationary process, so is that process stationary? For to be stationary, models statistical parameters(mean,covariance) should not be depend on time, so in my case let us suppose that amplitude, frequency and phase are some numbers,constants (they are not uniformly randomly distributed), so please help me to determine if this process is stationary and how good is to model this process by autoregressive model?

$\endgroup$
  • $\begingroup$ i found answer,it is not stationary if all parameters are fixed $\endgroup$ – dato datuashvili Apr 11 '14 at 8:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.