Let $(\Omega, \mathcal{F}, \mathbb{P})$ be a probability space and $( \mathcal{X}, \mathcal{B})$ be a measurable space. Let $\{X_t\}_{t\in [0,1]}$ be an uncountable collection of random variables such that $$X_t:(\Omega, \mathcal{F}, \mathbb{P}) \to ( \mathcal{X}, \mathcal{B})$$

When does $\{X_t\}$ define a stochastic process? For instance, if $X_t$ are iid standard normal random variables, then the corresponding collection doesn't define a stochastic process, (though I am not sure why).


This is an exercise from Lectures on Gaussian processes by Lifshits

Ex 1.3: Let $\mathcal{X}$ be an infinite dimensional separable Hilbert space and $E:\mathcal{X} \to \mathcal{X}$ be the identity operator. Prove that the distribution $\mathcal{N}(0, E)$ does not exist. Hint: If $X$ is a random vector with distribution $\mathcal{N}(0, E)$, it would satisfy an absurd identity $\mathbb{P}(||X||^2 = \infty) = 1$.

Here $\mathcal{N}(0, E)$ refers to the Gaussian process with identity covariance kernel and $0$ mean.

  • Why do you think this collection doesn't define a stochastic process? – Did Jul 22 '15 at 14:25
  • Hi, I have added the relevant example from the book. It does not claim that $\{X_t\}$ does not define a stochastic process but that the distribution does not exist. – user2808118 Jul 22 '15 at 14:31
  • Does this mean that a stochastic process can exist even if the corresponding distribution does not exist? – user2808118 Jul 22 '15 at 14:38
  • A white noise process you initially asked about does not have distribution $\mathcal{N}(0, E)$ as defined in Lifshits. As the exercise asks you to show, no random vector with values from $\mathcal{X}$ exists that would have the property that $(f,X)$ is normally distributed for every functional $f$. – Zoran Loncarevic Jul 22 '15 at 14:52
  • Okay. That answers the question. Thanks :) – user2808118 Jul 22 '15 at 14:55
up vote 1 down vote accepted

Stochastic process is a collection of random variables — nothing more and nothing less. That means your collection of iid standard normal random variables is an example of stochastic process that even has a special name — it's called a white noise.

  • Hi, I have added the relevant example from the book. It does not claim that $\{X_t\}$ does not define a stochastic process but that the distribution does not exist. – user2808118 Jul 22 '15 at 14:31

The only requirement is that the random variables $X_t$ are defined on the same probability space.

  • Hi, I have added the relevant example from the book. It does not claim that $\{X_t\}$ does not define a stochastic process but that the distribution does not exist. – user2808118 Jul 22 '15 at 14:31

Your Answer

 

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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