⟨Ax ∣ y⟩ = ⟨x ∣ Ay⟩
A linear operator A : X → Y is called bounded if it maps bounded subsets to bounded subsets.
A linear map between Banach spaces is continuous if and only if its graph is closed.
Makes proving continuity easier; you never have to show convergence, you just need to prove you converge to the correct thing.
Let A be a symmetric operator defined everywhere on a Hilbert space. Then A is bounded.
Let A be a linear operator from a bornological space to a locally convex space. Then A is bounded if and only if it's continuous.
Counter-example: Weak Topology
Every commutative unital C*-algebra A is isometrically *-isomorphic to C(K), the space of continuous complex-valued functions on a compact Hausdorff space K
Every arbitrary C*-algebra A admits a faithful, isometric *-representation as a closed subalgebra of B(H), the algebra of bounded operators on some Hilbert space H.
For any von Neumann algebra in standard form with a cyclic and separating vector, there exists a canonical one-parameter quantum evolution group (modular automorphism group).
Let p : V → ℝ be a seminorm on a real or complex vector space. Given any linear functional ϕ : U → k on a subspace U ⊆ V such that |ϕ| ≤ p on U, there exists a linear functional Φ : V → k such that Φ = ϕ on U and |Φ| ≤ p on V.
It's actually quite easy to find bounded linear functionals: you just need to ch Bounding a linear functional on a subspace is enough to be sure that it can be
It assures you that there are enough continuous linear functionals, but that's sort of an abstract reason.
Concretely there are 3 use cases:
Separation of Points
If you have two disjoint convex sets, you can find a closed hyperplane that separates them
Existence Proofs
Consider a problem like finding the shortest distance from a point x to a closed subspace C.
You can find a linear functional such that ϕ(c) = 0 for all c ∈ C ϕ(x) = 1 ||ϕ|| = 1/d
Embeddings
Given any Banach space X and x ≠ 0 we can find some functional with ||ϕ|| = 1 and ϕ(x) = ||x||
From here, we can embed X into X″ (the double continuous dual) ϕ : X → X″ with ϕx(f) = f(x)
We need to show that ||ϕx|| = ||x|| Now ||ϕx|| = sup||f|| ≤ 1|f(x)|
||f(x)|| ≤ ||f||||x|| ≤ ||x|| for ||f|| ≤ 1 And so ||ϕx|| ≤ ||x||. For the other direction, we use the HB theorem:
We look at the subspace generated by x and let ψ(λx) = λ||x|| Its norm is 1 Use HB to extend this to the full space X, and we get a linear functional ||Ψ|| ≤ 1 with Ψ(x) = ||x|| So ||ϕx|| = sup||f|| ≤ 1|f(x)| ≥ Ψ(x) = ||x||
Generalized Limits
You can define limit functionals on all convergent sequences, and extend that to all bounded sequences.