Proving Density of U and V in Hilbert Space: Tips and Tricks

Click For Summary
SUMMARY

The discussion centers on proving that the direct sum of the closed vector subspaces U and V is dense in the Hilbert space \(\ell_2^\infty\). U is generated by the basis vectors \{ e_{2k-1} | k \geq 1 \}, while V is generated by \{ e_{2k-1} + (1/k)e_{2k} | k \geq 1 \}. The user attempts various proof strategies, including set theoretic methods and projection techniques, ultimately seeking a method to demonstrate that for any point in \(\ell_2^\infty\), there exists a point in U \oplus V within an epsilon distance. The discussion highlights the need for a constructive approach to show density.

PREREQUISITES
  • Understanding of Hilbert spaces, specifically \(\ell_2^\infty\)
  • Familiarity with vector subspaces and direct sums
  • Knowledge of convergence in the context of sequences
  • Basic principles of linear algebra and functional analysis
NEXT STEPS
  • Study the properties of closed vector subspaces in Hilbert spaces
  • Learn about the concept of density in functional analysis
  • Explore projection theorems in Hilbert spaces
  • Investigate the use of epsilon-delta arguments in proofs of convergence
USEFUL FOR

Mathematicians, students of functional analysis, and anyone studying the properties of Hilbert spaces and vector subspaces will benefit from this discussion.

noospace
Messages
69
Reaction score
0

Homework Statement



Let [itex]e_i = (0,0,\ldots, 1, 0 , \ldots)[/itex] be the basis vectors of the Hilbert space [itex]\ell_2^\infty[/itex].

Let U and V be the closed vector subspaces generated by [itex]\{ e_{2k-1}|k \geq 1 \}[/itex] and [itex]\{ e_{2k-1} + (1/k)e_{2k} | k \geq 1 \}][/itex].

Show [itex]U \oplus V[/itex] dense in [itex]\ell_2^\infty[/itex]

I am looking for hints that anyone can offer.

The Attempt at a Solution



My main problem seems to be finding the most direct method of proof.

First I tried proving that [itex]\overline{U\oplus V} = \ell_2^\infty[/itex] using the set theoretic procedure of show each side of the equation is a subset or equal to the opposite side. I tried to use the fact that every element of the LHS has a sequence in U + V converging towards it. This got me nowhere.

I next tried showing that if x is any point in the hilbert space, there should be a point in the direct sum within a distance epsilon > 0 of it. I managed to construct an expression for x in terms of the basis vectors, but this expression is precisely x so is useless.

Currently I'm thinking about projecting x onto the subspaces and then using the sum of the projections somehow. It seems like this won't work however, since there no dependence on epsilon.

Any help would be greatly appreciated.

Thanks,.
 
Physics news on Phys.org
If (xn) is in l2, then given [itex]\epsilon[/itex] > 0, there exists N such that

[tex]\sum _{n > 2N}|x_n|^2 < \epsilon[/tex]

So at this point it suffices to show that for any N, (x1, x2, ..., x2N-1, x2N, 0, 0, ...) is in [itex]U \oplus V[/itex].
 

Similar threads

  • · Replies 43 ·
2
Replies
43
Views
5K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 26 ·
Replies
26
Views
2K
  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 6 ·
Replies
6
Views
4K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 4 ·
Replies
4
Views
2K