1

There're many versions of Riesz representation theorems saying that certain functionals $T$ on certain spaces $X$ can be represented by integrals of the form

$$ Tf = \int_X fd\mu $$

Where $\mu$ is some measure with certain features. For example Haar measures for locally compact groups, regular measures when dealing with $C(X)$, or $fd\mu$ with $f \in L^q(X,\mu)$ when the vector space is $L^p(X,\mu)$ (with $q$ conjugate exponent to $p$).

A question I have, for which I cannot find an answer is whether or not it is possible to characterize such measures by value on some "basis".

In finite dimensions a linear transformation is fully defined by the images over a basis for example. Basis for vector spaces of infinite dimensions are another story... I don't think it's that straightforward.

I am interested to know if $\sigma$-algebras have some notion of basis, just like topological spaces, and maybe there's some uniquiness theorem characterizing a measure over such basis.

I do know if we have a semi-algebra we can define the algebra generated by such sigma algebra and given an algebra we can define the sigma algebra generated by such algebra, I don't know if this is sufficient to characterize completely a both a sigma algebra and the measures over such algebra.

user8469759
  • 5,285
  • Perhaps you would be interested in the uniqueness part of Caratheodory Extension Theorem: https://en.wikipedia.org/wiki/Carath%C3%A9odory%27s_extension_theorem – geetha290krm Aug 18 '22 at 09:59
  • Yeah good point. Although I think I am more interested into knowing if there's a notion of "basis" for a $\sigma$-algebra, since I guess it would make easier to characterize measures. – user8469759 Aug 18 '22 at 10:54
  • Spanning sets corresponds to genearting sets but there is no notion of independence. – geetha290krm Aug 18 '22 at 11:25
  • Are you looking for this? https://math.stackexchange.com/questions/269011/is-there-a-similar-concept-for-a-sigma-algebra-like-a-base-for-a-topology – Yanko Aug 18 '22 at 12:18
  • Maybe, I am not sure. In topology given a basis for a topology you can generate a topological space and I think in Munkres Topology 2nd edition it is also proven the other way around, namely given a topological space you can somewhat find a basis. The semi-algebra allows me to generate a $\sigma$-algebra, but I don't know the other way around (i.e. given a $\sigma$-algebra is there a semi algebra who generates it?). – user8469759 Aug 18 '22 at 12:36

0 Answers0