2
$\begingroup$

Let $\mathfrak{M}$ be the sigma-algebra of the n-dimensional Lebesgue measurable sets (Completion of translation invariant measure having 1 for $[0,1]^n$)

Let $\Sigma$ be the sigma-algebra of the Lebesgue-measurable sets in $\mathbb{R}$.

Is $\mathfrak{M}$ the sigma-algebra generated by $\{\prod_{i=1}^n p(i) : p\in \Sigma^n\}$?

$\endgroup$
  • 1
    $\begingroup$ Almost. It's the completion of that $\sigma$-algebra with respect to the Lebesgue measure. $\endgroup$ – Daniel Fischer Feb 17 '14 at 19:24
  • $\begingroup$ @Daniel What do you mean by Almost? The sigma algebra generated by $\otimes_{i=1}^n \Sigma$ is bigger than the Borel algebra of $\mathbb{R}^n$ and i don't know whether it is smaller than the sigma-algenra of n-dim Lebesgue measurable sets. Nonetheless, of course its completion is the Sigma-algenra of n-dim Lebesgue measurable sets. I'm not asking that. Is it strictly smaller or the same? $\endgroup$ – John. p Feb 17 '14 at 19:44
  • $\begingroup$ @Daniel Note that $\otimes_{i=1}^n \Sigma$ is strictly bigger than the borel algebra of $\mathbb{R}^n$. Of course its completion is the sigma-algebra of n-dim Lebesgue measure. But I don't think it's easy to answer whether $\otimes_{i=1}^n \Sigma$ is exactly the sigma-algebra of n-dim Lebesgue measurable sets. And i believe it is. $\endgroup$ – John. p Feb 17 '14 at 19:51
5
$\begingroup$

The product $\sigma$-algebra $\bigotimes^n\Sigma$ lies strictly between the Borel $\sigma$-algebra $\mathcal{B}(\mathbb{R}^n)$ and the $\sigma$-algebra $\mathfrak{M}$ of Lebesgue-measurable subsets of $\mathbb{R}^n$ (for $n > 1$). Since the latter is already the completion of the Borel $\sigma$-algebra with respect to the Lebesgue measure, it is also the completion of $\bigotimes^n\Sigma$ with respect to the Lebesgue measure.

Since there are Lebesgue-measurable sets that are not Borel sets, the one strict containment is clear.

For the other, let $N\subset \mathbb{R}$ a non-measurable set. Then $N\times \{0\}^{n-1}$ is a Lebesgue null-set in $\mathbb{R}^n$, but does not belong to $\bigotimes^n\Sigma$. To see the latter, note that for $M\in \bigotimes^n\Sigma$, every section $M_y = \{x \in \mathbb{R} : (x,y) \in M\}, \; y \in \mathbb{R}^{n-1}$ arbitrary, lies in $\Sigma$.

$\endgroup$
  • 1
    $\begingroup$ Nice! Thank you so much! $\endgroup$ – John. p Feb 17 '14 at 20:01

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.