A Lognormal Central Limit Theorem for Particle Approximations of Normalizing Constants
نویسندگان
چکیده
Feynman-Kac models arise in a large variety of scientific disciplines including physics, chemistry and signal processing. Their mean field particle interpretations, termed commonly Sequential Monte Carlo or Particle Filters, have found numerous applications as they allow to sample approximately from sequences of complex probability distributions and estimate their associated normalizing constants. It is well-known that, under regularity assumptions, the relative variance of these normalizing constant estimates increases linearly with the time horizon n so that practitioners usually scale the number of particles N linearly w.r.t n to obtain estimates whose relative variance remains uniformly bounded w.r.t n. We establish here that, under this standard linear scaling strategy, the fluctuations of the normalizing constant estimates are lognormal as n, hence N , goes to infinity. For particle absorption models in a time-homogeneous environment and hidden Markov models in an ergodic random environment, we also provide more explicit descriptions of the limiting bias and variance.
منابع مشابه
The Constants in the Clt for the Edwards Model
The Edwards model in one dimension is a transformed path measure for one-dimensional Brownian motion discouraging self-intersections. In van der Hofstad, den Hollander and KK onig (preprint 1995) a central limit theorem (CLT) is proved for the uctuations of the endpoint of the path around its linear asymptotics. In the present paper, we study the constants appearing in this CLT (which represent...
متن کاملCentral Limit Theorem in Multitype Branching Random Walk
A discrete time multitype (p-type) branching random walk on the real line R is considered. The positions of the j-type individuals in the n-th generation form a point process. The asymptotic behavior of these point processes, when the generation size tends to infinity, is studied. The central limit theorem is proved.
متن کاملThe Local Limit Theorem: A Historical Perspective
The local limit theorem describes how the density of a sum of random variables follows the normal curve. However the local limit theorem is often seen as a curiosity of no particular importance when compared with the central limit theorem. Nevertheless the local limit theorem came first and is in fact associated with the foundation of probability theory by Blaise Pascal and Pierre de Fer...
متن کاملApproximations on the Aggregate MPEG Video Traffic and Their Impact on Admission Control
In this paper, we investigate the aggregate traffic approximations for the Motion Pictures Expert Group (MPEG) coded variable bit rate (VBR) video traffic and their impact on admission control at a desired Quality of Service (QoS) level. In order to generate the MPEG coded VBR video traffic, we use a novel source model that employs a mixture of two first-order autoregressive processes with logn...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2014