Feynman–Kac conjecture for the stochastic Bessel operator

About 9 years old · traced to

Let B\mathcal{B} be the stochastic Bessel operator, let f∈L2[0,1]f\in L^2[0,1], and let XX be the stochastic process formally satisfying

dX=2X dB+(1+2X/β W′(X)) dt,dX=\sqrt{2X}\,dB+\left(1+2\sqrt{X/\beta}\,W'(X)\right)\,dt,

with functional

Φ(X)=−a24∫0tduXu−aβ∫01LX(x,t)x∘dW(x).\Phi(X)=-\frac{a^2}{4}\int_0^t\frac{du}{X_u}-\frac{a}{\sqrt{\beta}}\int_0^1\frac{L_X(x,t)}{\sqrt{x}}\circ dW(x).

Let AA be the event sup⁡0≤u≤tXu≤1\sup_{0\leq u\leq t}X_u\leq 1. Feynman–Kac conjecture. The stochastic Bessel operator satisfies

exp⁡(−tB)f(x)=Ex[1Aexp⁡Φ(X)f(Xt)].\exp(-t\mathcal{B})f(x)=\mathbb{E}_x\left[\mathbb{1}_A\exp\Phi(X)f(X_t)\right].

Moreover, if fnf_n is the orthogonal projection of a fixed f∈L2[0,1]f\in L^2[0,1] onto the span of {1[(k−1)/n,k/n):k=1,…,n}\{\mathbb{1}_{[(k-1)/n,k/n)}:k=1,\ldots,n\}, then (I−L)tn2fn(I-L)^{tn^2}f_n should converge in distribution to the right-hand side of this formula, where LL is a matrix from the n×nn\times n tridiagonal β\beta-Laguerre ensemble. This conjecture seeks to identify the stochastic Bessel operator semigroup through a continuum Feynman–Kac representation and to connect it with large powers of the finite-dimensional tridiagonal Laguerre model; the stated convergence is the proposed, rather than established, link.

References

Primary source

Patrick Waters, “Feynman-Kac formula for the stochastic Bessel operator”, arXiv:1711.00908 (2017).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.