A probabilistic solution to the Stroock-Williams equation
Goran Peskir

TL;DR
This paper provides a probabilistic representation for solutions to a boundary value problem related to the Stroock-Williams equation, involving reflecting Brownian motion with drift and local time, even when boundary conditions are not of Feller's type.
Contribution
It introduces a novel probabilistic formula for the solution of the Stroock-Williams equation with non-Feller boundary conditions, using reflecting Brownian motion and local time.
Findings
Valid probabilistic representation of the solution.
Closed integral formula involving drift and boundary parameters.
Applicable to boundary conditions not of Feller's type.
Abstract
We consider the initial boundary value problem \begin{eqnarray*}u_t=\mu u_x+\tfrac{1}{2}u_{xx}\qquad (t>0,x\ge0),\\u(0,x)=f(x)\qquad (x\ge0),\\u_t(t,0)=\nu u_x(t,0)\qquad (t>0)\end{eqnarray*} of Stroock and Williams [Comm. Pure Appl. Math. 58 (2005) 1116-1148] where and the boundary condition is not of Feller's type when . We show that when belongs to with then the following probabilistic representation of the solution is valid: \[u(t,x)=\mathsf{E}_x\bigl[f(X_t)\bigr]-\mathsf{E}_x\biggl[f'(X_t)\int_0^{\ell_t^0(X)}e^{-2(\nu-\mu)s}\,ds\biggr],\] where is a reflecting Brownian motion with drift and is the local time of at . The solution can be interpreted in terms of and its creation in at rate proportional to . Invoking the law of , this also yields a closed integral…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
