Problem of the Week #107 - June 16th, 2014

  • Thread starter Thread starter Chris L T521
  • Start date Start date
Click For Summary
SUMMARY

This discussion addresses the continuity of linear transformations between normed linear spaces. It establishes that if the domain space $X$ is finite dimensional, then any linear transformation $T: X \rightarrow Y$ is continuous. Conversely, if the codomain space $Y$ is finite dimensional, $T$ is continuous if and only if the kernel of $T$ is closed. The proofs utilize properties of finite dimensional spaces, including the equivalence of norms and the behavior of sequences in relation to continuity.

PREREQUISITES
  • Understanding of linear transformations in normed linear spaces
  • Familiarity with the concepts of continuity and closed sets in topology
  • Knowledge of finite dimensional vector spaces and their properties
  • Proficiency in working with norms, particularly the 1-norm and its equivalence in finite dimensions
NEXT STEPS
  • Study the properties of linear transformations in finite dimensional spaces
  • Learn about the implications of the closed graph theorem in functional analysis
  • Explore the relationship between continuity and closed sets in normed spaces
  • Investigate the equivalence of norms in finite dimensional vector spaces
USEFUL FOR

Mathematicians, students of functional analysis, and anyone studying linear algebra who seeks to understand the continuity of linear transformations in normed spaces.

Chris L T521
Gold Member
MHB
Messages
913
Reaction score
0
Here's this week's problem!

-----

Problem
: For normed linear spaces $X$ and $Y$, let $T:X\rightarrow Y$ be linear. If $X$ is finite dimensional, show that $T$ is continuous. If $Y$ is finite dimensional, show that $T$ is continuous if and only if $\ker T$ is closed.

-----

Remember to read the http://www.mathhelpboards.com/showthread.php?772-Problem-of-the-Week-%28POTW%29-Procedure-and-Guidelines to find out how to http://www.mathhelpboards.com/forms.php?do=form&fid=2!
 
Physics news on Phys.org
This week's problem was correctly answered by Euge. You can find his solution below.

[sp]Suppose $X$ is finite dimensional. Let {$e_1, e_2, \ldots, e_m$} be a basis for $X$. Given $x \in X$, there are unique scalars $x^1, x^2,\ldots, x^m$ such that $x = \sum_{i = 1}^m x^i e_i$. By linearity of $T$ and the triangle inequality,

$\|Tx\|_Y = \left\|\sum_{i = 1}^m x^i Te_i\right\|_Y \le \sum_{i = 1}^m |x^i| \|Te_i\|_Y \le c \|x\|_1$,

where $c = \max\{\|Te_1\|_Y, \|Te_2\|_Y,\ldots, \|Te_m\|_Y\}$ and $\| \cdot \|_1$ is the 1-norm with respect to the given basis for $X$. Since $X$ is finite dimensional, $\| \cdot \|_1$ is equivalent to $\| \cdot \|_X$. So $\|x\|_1 \le d\|x\|_X$ for some constant $d$ independent of $x$. Hence $\|Tx\|_Y \le cd \|x\|_X$. Since $x$ was arbitrary and constants $c$ and $d$ are independent of $x$, it follows that $T$ is continuous.

Next, suppose $Y$ is finite dimensional. Assume $T$ is continuous. For a given $x$ in the closure of ker $T$, there exists a sequence $(x_n)$ in $X$ such that $\lim_n x_n = x$. Since $Tx_n = 0$ for all $n$, continuity of $T$ yields $Tx = \lim_n Tx_n = 0$, i.e., $x\in \text{ker} T$. Thus ker $T$ is closed.

Conversely, suppose ker $T$ is closed. Then the quotient space $X/\text{ker}\, T$ is a normed linear space with norm $\|x + \text{ker}\, T\| := \inf\{\|x - u\|: u \in \text{ker} T\}$. As a vector space, $X/\text{ker}\, T$ is isomorphic to a subspace of the finite dimensional vector space $Y$, so $X/\text{ker}\, T$ is finite dimensional. Furthermore, linearity of $T$ implies linearity of the induced map $X/\text{ker}\, T \to Y$. So by the argument given in the first paragraph, the induced map is continuous. The natural projection map $\pi : X \to X/\text{ker}\, T$ is continuous since $\|\pi(x)\| \le \|x\|_X$ for all $x \in X$. Since the composition of continuous maps is continuous and $T$ is the composition $X \to X/\text{ker}\, T \to Y$, it follows that $T$ is continuous.[/sp]
 

Similar threads

  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 1 ·
Replies
1
Views
1K