The weak* topology on dual spaces is a crucial concept in functional analysis. It's defined on the dual space X* using seminorms and allows for a more relaxed notion of convergence compared to the norm topology.

Weak* topology is coarser than the weak topology on X*, making it easier to work with in certain contexts. The Banach-Alaoglu theorem, which states that the closed unit ball in X* is weak* compact, is a key result in this area.

The Weak Topology on Dual Spaces

Weak topology in dual spaces

Defined on the dual space $X^*$ $X^{*}$ of a normed space $X$ $X$ using the family of seminorms $\{p_x : x \in X\}$ ${p_{x} : x \in X}$
- Each seminorm $p_x(f)$ evaluates the absolute value of the functional $f$ at the point $x$ , i.e., $p_x(f) = |f(x)|$
Subbase for the weak* topology consists of sets $\{f \in X^* : |f(x) - f_0(x)| < \varepsilon\}$ ${f \in X^{*} : ∣ f (x) - f_{0} (x) ∣ < ε}$
- Determined by a point $x \in X$ , a functional $f_0 \in X^*$ , and a positive real number $\varepsilon > 0$
Convergence in the weak* topology is equivalent to pointwise convergence on $X$ $X$
- A net $(f_\alpha)$ in $X^*$ converges to $f \in X^*$ in the weak* topology if and only if $f_\alpha(x) \to f(x)$ for each $x \in X$

Weak vs weak topology comparison

Weak topology on $X^*$ defined by seminorms $\{p_x : x \in X\}$ with $p_x(f) = \|f\| \|x\|$
Every subbase element of the weak* topology is also a subbase element of the weak topology
- For $x \in X$ $x \in X$ , $f_0 \in X^*$ $f_{0} \in X^{*}$ , $\varepsilon > 0$ $ε > 0$ , the set $\{f \in X^* : |f(x) - f_0(x)| < \varepsilon\}$ ${f \in X^{*} : ∣ f (x) - f_{0} (x) ∣ < ε}$ is open in the weak topology
  - Follows from the inequality $|f(x) - f_0(x)| \leq \|f - f_0\| \|x\|$
The weak* topology is coarser than the weak topology on $X^*$ $X^{*}$
- Every weak* open set is also weakly open, but not conversely

Weak* topology in dual spaces, Frontiers | Topological Schemas of Memory Spaces

Weak Closed and Compact Sets in Dual Spaces

Characteristics of weak sets

A subset $A \subset X^*$ $A \subset X^{*}$ is weak* closed if and only if it is closed under pointwise limits
- Equivalent to: for every net $(f_\alpha)$ in $A$ that converges to $f \in X^*$ in the weak* topology, $f \in A$
Banach-Alaoglu theorem: the closed unit ball $B_{X^*} = \{f \in X^* : \|f\| \leq 1\}$ is compact in the weak* topology
A subset $A \subset X^*$ is weak* compact if and only if it is weak* closed and bounded in the norm topology
Weak* compact sets in $X^*$ $X^{*}$ have the following properties:
- Every net in $A$ has a subnet that converges in the weak* topology to an element of $A$
- $A$ is compact in the weak* topology if and only if every continuous linear functional on $(X^*, \text{weak*})$ attains its maximum on $A$

Relationship of weak and weak topologies

If $X$ $X$ is reflexive (canonical embedding $J : X \to X^{**}$ $J : X \to X^{**}$ is surjective), the weak* and weak topologies on $X^*$ $X^{*}$ coincide
- In this case, weak* closed (resp. compact) sets are the same as weakly closed (resp. compact) sets
If $X$ $X$ is not reflexive, the weak* topology is strictly coarser than the weak topology on $X^*$ $X^{*}$
- There exist subsets of $X^*$ $X^{*}$ that are:
  - Weak* closed but not weakly closed
  - Weak* compact but not weakly compact

2,589 studying →