What is the Maximum Norm Proof for Matrix A?

In summary, we are trying to prove that for a square matrix $A$, the infinity norm $\|A\|_\infty$ is equal to the maximum of the sums of absolute values of the rows of $A$. We can use the equivalent definition of $\|A\|_\infty$ as the maximum of $\|Ax\|_\infty$ over all unit vectors $x$. To prove this, we can show that for any unit vector $x$, $\|Ax\|_\infty$ is less than or equal to the sum of the absolute values of the $i$th row of $A$, and for a specific choice of $x$ we can show that $\|Ax\|_\in
  • #1
Amer
259
0
Prove that for

$A \in \mathbb{R}^{n\times n} $
[tex]||A||_{\infty} = \text{max}_{i=1,...,n} \sum_{j=1}^n |a_{ij} | [/tex]

I know that
$||A||_{\infty} = \text{max} \dfrac{||Ax||_{\infty} }{||x||_{\infty}} $

such that $x \in \mathbb{R}^n$

any hints
 
Physics news on Phys.org
  • #2
Amer said:
Prove that for

$A \in \mathbb{R}^{n\times n} $
[tex]||A||_{\infty} = \text{max}_{i=1,...,n} \sum_{j=1}^n |a_{ij} | [/tex]

I know that
$||A||_{\infty} = \text{max} \dfrac{||Ax||_{\infty} }{||x||_{\infty}} $

such that $x \in \mathbb{R}^n$

any hints

I believe you can start with $||A||_p$ the p norm and take the limit as $p\to\infty$ to prove the problem.
Isn't the infinity norm just the max of $|a_i|$ not the sum of them?
 
  • #3
dwsmith said:
I believe you can start with $||A||_p$ the p norm and take the limit as $p\to\infty$ to prove the problem.
Isn't the infinity norm just the max of $|a_i|$ not the sum of them?
for a vector it is the max of $|a_i|$
 
  • #4
Amer said:
Prove that for

$A \in \mathbb{R}^{n\times n} $
[tex]||A||_{\infty} = \text{max}_{i=1,...,n} \sum_{j=1}^n |a_{ij} | [/tex]

I know that
$||A||_{\infty} = \text{max} \dfrac{||Ax||_{\infty} }{||x||_{\infty}} $

such that $x \in \mathbb{R}^n$

any hints
You might find it easier to use the equivalent definition $\|A\|_\infty = \max \{\|Ax\|_\infty : \|x\|_\infty \leqslant 1\}.$ For $\|x\|_\infty \leqslant 1$, show that $$\|Ax\|_\infty = \max_{1\leqslant i\leqslant n}|(Ax)_i| = \max_{1\leqslant i\leqslant n}\Bigl| \sum_{j=1}^n a_{ij}x_j \Bigr| \leqslant \sum_{j=1}^n |a_{ij} |.$$
For the reverse inequality, find $\|Ax\|$ where $x$ is the vector with a 1 in the $i$th coordinate and 0 for every other coordinate.
 
  • #5


The Maximum Norm Proof for Matrix A states that the maximum norm of a matrix A, denoted as ||A||_∞, is equal to the maximum absolute row sum of A. In other words, ||A||_∞ is equal to the maximum of the sums of the absolute values of the elements in each row of A.

To prove this, we can start by considering the definition of the maximum norm for a matrix A. This is defined as the maximum of the norms of all possible vectors x in ℝ^n, such that ||x||_∞ = 1. In other words, we are looking for the maximum value of ||Ax||_∞ over all possible unit vectors x.

We can rewrite ||Ax||_∞ as the maximum absolute row sum of A multiplied by the norm of x. This is because the norm of x is equal to 1, and the maximum absolute row sum of A is the maximum value that can be obtained by multiplying A with any unit vector x.

So, we have ||Ax||_∞ = ||A||_∞ ||x||_∞. Substituting this into the original definition, we get:

||A||_∞ = max ||Ax||_∞ / ||x||_∞

Now, let's consider the definition of the maximum absolute row sum of A. This is equal to the maximum of the sums of the absolute values of the elements in each row of A. We can rewrite this as:

max ||Ax||_∞ / ||x||_∞ = max (|a_{11}| + |a_{12}| + ... + |a_{1n}|, |a_{21}| + |a_{22}| + ... + |a_{2n}|, ..., |a_{n1}| + |a_{n2}| + ... + |a_{nn}|)

Since we are looking for the maximum value of this expression, we can simplify it further by considering the maximum value of each term in the expression. This gives us:

max ||Ax||_∞ / ||x||_∞ = max (max |a_{11}|, max |a_{12}|, ..., max |a_{1n}|, max |a_{21}|, ..., max |a_{n1}|, ..., max |a_{nn}|)

Now, we know that the maximum value of any term in
 

FAQ: What is the Maximum Norm Proof for Matrix A?

What is a "Maximum Norm Proof"?

A Maximum Norm Proof is a mathematical proof technique used to show that a function or sequence converges to a maximum value. It is often used in analysis and optimization problems.

How is a "Maximum Norm Proof" different from other proof techniques?

Unlike other proof techniques, such as limit proofs or delta-epsilon proofs, a Maximum Norm Proof focuses specifically on convergence to a maximum value rather than a limit. It is also often used in the context of optimization problems, rather than general mathematical functions.

What are the main steps involved in a "Maximum Norm Proof"?

The main steps of a Maximum Norm Proof include defining the function or sequence, showing that it is bounded above by the maximum value, and then showing that it converges to the maximum value. This is often done using the definition of a limit and the properties of the maximum norm.

When is a "Maximum Norm Proof" useful?

A Maximum Norm Proof is useful in situations where it is necessary to show that a function or sequence converges to a maximum value. This is often the case in optimization problems, where the goal is to find the maximum value of a given function.

Are there any limitations to using a "Maximum Norm Proof"?

One limitation of a Maximum Norm Proof is that it can only be used to show convergence to a maximum value. It cannot be used to show convergence to a minimum value or any other specific value. Additionally, it may not be the most efficient proof technique in all situations, and other methods may be more appropriate.

Similar threads

Replies
19
Views
4K
Replies
12
Views
2K
Replies
1
Views
1K
Replies
8
Views
2K
Replies
1
Views
2K
Replies
52
Views
3K
Replies
5
Views
2K
Back
Top