0
$\begingroup$

If I understand correctly, a homogeneous process must be homogeneous-increment, because for example, $L(X_{t_2} - X_{t_1}) = L(X_{t_2 + \tau} - X_{t_1 + \tau})$ can be proven by $$L(X_{t_2} - X_{t_1}, X_{t_1}) = L(X_{t_2}, X_{t_1}) = L(X_{t_2 + \tau}, X_{t_1 + \tau}) = L(X_{t_2 + \tau} - X_{t_1 + \tau}, X_{t_1 + \tau})$$ and then marginalize $L(X_{t_2} - X_{t_1}, X_{t_1})$ wrt $X_{t_1}$, and $L(X_{t_2 + \tau} - X_{t_1 + \tau}, X_{t_1 + \tau})$ wrt $ X_{t_1 + \tau}$ and the equality still holds.

I am now wondering about the reverse.

  1. What are some homogeneous-increment process that are not homogeneous?

  2. Is it possible to have some necessary and sufficient condition for a homogeneous-increment process to be homogeneous?

Thanks and regards!

$\endgroup$
4
$\begingroup$

What are some homogeneous-increment process that are not homogeneous?

$$X_t=t$$

$\endgroup$
  • $\begingroup$ +1 Thanks! Do you happen to know some necessary and/or sufficient condition for a homogeneous-increment process to be homogeneous? $\endgroup$ – Tim Dec 12 '12 at 15:45

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.