WorksMathematics
Exact Wasserstein Convergence Rate for Conditional Empirical Measures of Killed Diffusion Processes
Abstract
Let be a -dimensional compact connected Riemannian manifold with boundary \pp M, and let be the diffusion process generated by L:=\DD+\nn V for some which is killed on the boundary. For any initial distribution which is not fully supported on \pp M, let be the conditional empirical measure for the diffusion process under the condition that the process is not yet killed at time , and let be the associated quasi-stationary distribution. We derive the sharp convergence rate of to in the quadratic Wasserstein distance \W_2 as . In particular, if for some and all uniformly in , then for large \W_2(μ_t^ν,μ_0)\asymp_ν\begin{cases} t^{-1},&d-k<6,\\ t^{-1}\sqrt{\log t},&d-k=6,\\ t^{-\ff{d-k+2}{2(d-k-2)}},&d-k>6. \end{cases} Moreover, letting \rr_\pp(x) be the distance of to \pp M, for large we have \W_2(μ_t^{\dd_x},μ_0)\asymp \begin{cases} t^{-1},&d\le 3,\\ t^{-1}\log (1+\rr_\pp(x)^{-1}), & d=4,\\ t^{-1} \rr_\pp(x)^{-1}, & d=5,\\ t^{-1}\sqrt{\log t} \rr_\pp(x)^{-1},&d=6,\\ t^{-\ff{d+2}{2(d-2)}}\rr_\pp(x)^{-\ff 4{d-2}},&d\ge 7. \end{cases}