Positive definite operator/matrix question

  • Thread starter Thread starter holezch
  • Start date Start date
  • Tags Tags
    Positive
Click For Summary

Homework Help Overview

The discussion revolves around proving that a linear operator T is positive definite based on a specific summation involving a matrix A and a non-zero tuple. The context is within linear algebra, particularly focusing on properties of matrices and operators.

Discussion Character

  • Exploratory, Conceptual clarification, Mathematical reasoning

Approaches and Questions Raised

  • Participants explore the relationship between the summation and matrix multiplication, with one suggesting to consider the expression \(\overline{a}^t A a\). Others express confusion about the implications of these expressions and the conditions for T being positive definite.

Discussion Status

The discussion is ongoing, with participants sharing insights and questioning the significance of the mathematical expressions involved. Some guidance has been offered regarding the matrix representation, but there is no clear consensus on the next steps or interpretations of the problem.

Contextual Notes

Participants note potential confusion regarding the dimensionality and structure of the vectors and matrices involved, particularly in relation to the definitions of positive definiteness and the properties of the operator T.

holezch
Messages
251
Reaction score
0

Homework Statement



Prove that T is positive definite if and only if

[tex]\sum_{i,j} A_{ij}a_{j}\bar{a_{i}} > 0[/tex]
for any non-zero tuple (a1, ...... , an )

Let A be [tex][ T ]_{\beta}[/tex]

where [tex]\beta[/tex] is an orthogonal basis for T

The Attempt at a Solution



the sum looked like the matrix multiplication of a n-tuple and a matrix A, so I looked into that and couldn't get anything.. any hints please? I'm also struggle to realize what significant that sum could have, right now it doesn't even mean anything to me.

Thanks!
 
Physics news on Phys.org
If the capital A's are supposed to be terms in matrix, you might want to think about

[tex]\overline{a}^tAa[/tex]
 
Office_Shredder said:
If the capital A's are supposed to be terms in matrix, you might want to think about

[tex]\overline{a}^tAa[/tex]

thanks, someone else told me to think about [tex]a^{*} A a[/tex] , but I'm not sure why I'd be considering this. I basically have these things to work with: that if T is positive definite ( one way of the implication ), then all its eigenvalues are positive,T is self-adjoint and < Tx , x > > 0.

Thanks
 
I'm also confused a*A might not make sense.. if a is a column vector in F^n and A is a matrix?
 

Similar threads

Replies
5
Views
2K
  • · Replies 5 ·
Replies
5
Views
1K
  • · Replies 31 ·
2
Replies
31
Views
4K
  • · Replies 12 ·
Replies
12
Views
3K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 3 ·
Replies
3
Views
3K
Replies
5
Views
2K
  • · Replies 5 ·
Replies
5
Views
14K