Essential boundedness conjecture for geometric ergodicity of i-cSMC

From papers

Let TNT\in\mathbb{N}, let πt\pi_t denote the time-tt marginal of the target distribution, and let GtG_t be the corresponding potential function. Write PNP_N for the iterated conditional SMC (i-cSMC) kernel with NN particles. For a measurable function, πt\pi_t-esssupxtGt(xt){\rm ess}\sup_{x_t}G_t(x_t) denotes its essential supremum with respect to πt\pi_t. Essential boundedness conjecture. If

πt-esssupxtGt(xt)=\pi_t\text{-}{\rm ess}\sup_{x_t}G_t(x_t)=\infty

for some t[T]t\in[T], then the i-cSMC kernel is not geometrically ergodic for any NNN\in\mathbb{N}. This conjecture proposes necessity of essential boundedness of the potentials for geometric ergodicity. The paper notes that this conclusion holds in several examples through sticky sets, but does not establish the general claim; hence its status remains open.

Progress summary

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

Sources & referencesView supporting material

Primary source

Christophe Andrieu, Anthony Lee and Matti Vihola, “Uniform Ergodicity of the Iterated Conditional SMC and Geometric Ergodicity of Particle Gibbs samplers”, arXiv:1312.6432 (2015).

Solutions 0

No solutions have been posted yet.