TheoremBase

Comparison and Uniqueness under Wasserstein Semicontinuity for the Controlled von Neumann HJB Equation with Free Langevin Noise in a Wall, Including the Free Entropy Penalty

For the controlled von Neumann HJB equation with free Langevin noise in a wall and a uniformly continuous running cost, a bounded Wasserstein-upper-semicontinuous penalised subsolution lies below a bounded Wasserstein-lower-semicontinuous penalised supersolution on the wall domain, so such solutions are unique; this covers the free entropy penalty under the displacement convexity conjecture.

Statement

In the setting of The Discounted Hamilton-Jacobi-Bellman Equation with Free Langevin Noise in a Wall on Noncommutative Laws: Standing Notation, let ff and HvN\mathcal{H}^{\mathrm{vN}} be as in The Controlled von Neumann Hamilton-Jacobi-Bellman Equation with Free Langevin Noise in a Wall on Noncommutative Laws §data, let (EvN)(\mathrm{E}^{\mathrm{vN}}) be the equation of The Controlled von Neumann Hamilton-Jacobi-Bellman Equation with Free Langevin Noise in a Wall on Noncommutative Laws §equation, which is the equation of The Discounted Hamilton-Jacobi-Bellman Equation with Free Langevin Noise in a Wall on Noncommutative Laws §score-form with the Hamiltonian H=HvN\mathcal{H}=\mathcal{H}^{\mathrm{vN}}, let RR be the wall radius of The Discounted Hamilton-Jacobi-Bellman Equation with Free Langevin Noise in a Wall on Noncommutative Laws: Standing Notation §parameters, and let D\mathcal{D} be the domain of the wall-confined free energy as in The Discounted Hamilton-Jacobi-Bellman Equation with Free Langevin Noise in a Wall on Noncommutative Laws: Standing Notation §free-energy. Semicontinuity on D\mathcal{D} refers to the metric space (Σd,R,W2)(\Sigma_{d,R},W_{2}) of The Noncommutative Laws with a Norm Bound Form a Complete Bounded Metric Space with Interpolation Points, which contains D\mathcal{D} by The Wall-Confined Free Energy: Norm Bound, Lower Bound, Weak-Star Compact Sublevel Sets and Displacement Monotonicity of the Score §bounds. Assume that ff is uniformly continuous on Σd2\Sigma^{2}_{d} with the metric of Plan Jets and Hamiltonians on Square-Integrable Noncommutative Laws: Standing Notation §metrics.

1. (Comparison) Let u,v:Σd2→Ru,v:\Sigma^{2}_{d}\to\mathbb{R} be bounded; let uu be a free-energy-penalised viscosity subsolution of (EvN)(\mathrm{E}^{\mathrm{vN}}) such that the function μ↦u(κd(μ))\mu\mapsto u(\kappa_{d}(\mu)) on D\mathcal{D} is upper semicontinuous on D\mathcal{D}, and let vv be a free-energy-penalised viscosity supersolution of (EvN)(\mathrm{E}^{\mathrm{vN}}) such that the function μ↦v(κd(μ))\mu\mapsto v(\kappa_{d}(\mu)) on D\mathcal{D} is lower semicontinuous on D\mathcal{D}. Then u(κd(μ))≤v(κd(μ))u(\kappa_{d}(\mu))\le v(\kappa_{d}(\mu)) for every μ∈D\mu\in\mathcal{D}.

2. (Uniqueness) Any two bounded free-energy-penalised viscosity solutions u,v:Σd2→Ru,v:\Sigma^{2}_{d}\to\mathbb{R} of (EvN)(\mathrm{E}^{\mathrm{vN}}) such that the functions μ↦u(κd(μ))\mu\mapsto u(\kappa_{d}(\mu)) and μ↦v(κd(μ))\mu\mapsto v(\kappa_{d}(\mu)) on D\mathcal{D} are both upper semicontinuous and lower semicontinuous on D\mathcal{D} agree at κd(μ)\kappa_{d}(\mu) for every μ∈D\mu\in\mathcal{D}.

3. (Free entropy penalty) Let Dχ\mathcal{D}_{\chi} and χ∗\chi^{*} be the domain and the free entropy of Free Products, Semicircular Laws, the Free Heat Flow, Free Fisher Information and Free Entropy: Standing Notation §entropy, let −χ∗:Dχ→R-\chi^{*}:\mathcal{D}_{\chi}\to\mathbb{R} be the map λ↦−χ∗(λ)\lambda\mapsto-\chi^{*}(\lambda), and assume that the pair (Dχ,−χ∗)(\mathcal{D}_{\chi},-\chi^{*}) satisfies the tangent inequality clause of a free entropy penalty, as conjectured in Displacement Convexity of Minus the Free Entropy: the Tangent Inequality along Optimal Couplings §tangent. Then (Dχ,−χ∗)(\mathcal{D}_{\chi},-\chi^{*}) is a free entropy penalty by Conditional Comparison and Uniqueness for the Linear-Quadratic HJB Equation with Free Langevin Noise in a Wall, Penalised by Voiculescu's Free Entropy §penalty, and consequently claims 1 and 2 apply to (D0,E0)=(Dχ,−χ∗)(\mathcal{D}_{0},\mathcal{E}_{0})=(\mathcal{D}_{\chi},-\chi^{*}), for which D=Dχ∩DR\mathcal{D}=\mathcal{D}_{\chi}\cap\mathcal{D}_{R}.

Proofs

Log in to submit a proof.

Loading...

Citations

Loading…

Dependencies

Loading…

Related

0 relations

Curated associations between results. These are editable and subjective — they do not replace the dependency graph, which is derived from the references in the text.

No relations recorded yet.

Comments

Log in to comment.

Loading…