5
$\begingroup$

Let us consider $\ell^1$, the space of absolutely summable sequences in the space of complex numbers. Clearly every finite dimensional Hilbert space is topologically embedded into $\ell^1$.

Convention. For given Hilbert spaces $H$ and $K$, let us write $H\leq K$ if the Hilbertian dimension of $H$ is less than $K$.

Q. What is the Hilbertian dimension of the largest Hilbert space which can be topologically embedded into $\ell^1$?

Improve this question
$\endgroup$
4
  • 1
    $\begingroup$ Your convention is not used in the question, so why include it? $\endgroup$
    – LSpice
    Commented May 9, 2018 at 18:46
  • $\begingroup$ Yes you are right. $\endgroup$
    – ABB
    Commented May 9, 2018 at 18:49
  • $\begingroup$ relevant: mathoverflow.net/questions/79713/… $\endgroup$
    – Pietro Majer
    Commented May 9, 2018 at 19:14
  • $\begingroup$ ANSWER: $\ l^2\ $ is TOPOLOGICALLY embeddable in $\ l^1$. $\endgroup$
    – Wlod AA
    Commented May 10, 2018 at 4:10

1 Answer 1

Reset to default
4
$\begingroup$

No infinite dimensional reflexive space can be embedded into $\ell_1$, because every infinite dimensional closed subspace of $\ell_1$ has a non separable dual.

Improve this answer
$\endgroup$
8
  • $\begingroup$ Since the question is just about a topological embedding, is it obvious that the image is closed? $\endgroup$
    – LSpice
    Commented May 9, 2018 at 18:54
  • $\begingroup$ @Pietro Majer What happens if we change the roles? I mean what is the Hilbertian dimension of the smallest Hilbert space contains $\ell^1$ up to topological vector spaces. $\endgroup$
    – ABB
    Commented May 9, 2018 at 18:55
  • 1
    $\begingroup$ @Spice yes because it is complete $\endgroup$
    – Pietro Majer
    Commented May 9, 2018 at 18:55
  • 1
    $\begingroup$ rmk: Of course if one only wants a homeo onto a subspace, not necessarily linear, then $\ell_2$ is already homeomorphic to $\ell_1$, via $\ell_2\ni x:=(x_j)_j\mapsto (|x_j|x_j)_j\in\ell_1$ $\endgroup$
    – Pietro Majer
    Commented May 9, 2018 at 19:03
  • 2
    $\begingroup$ @LSpice of course topological embedding (not necessarily linear) need not preserve completeness- e.g. R can be embedded in R as an open interval. But the question is only meaningful in the top. linear sense ($\ell_1$ and $\ell_2$ are already homeomorphic, as topological spaces) $\endgroup$
    – Pietro Majer
    Commented May 9, 2018 at 19:37

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .