Adaptive sequential Monte Carlo by means of mixture of experts
Adaptive sequential Monte Carlo by means of mixture of experts
复制标题
DOI：
10.1007/s11222-012-9372-2
复制
发表时间：
2011-08
影响因子：
2.2
通讯作者：
Julien Cornebise;É. Moulines;J. Olsson
中科院分区：
文献类型：
--
作者：
Julien Cornebise;É. Moulines;J. Olsson
适当地设计粒子滤波器的建议核是一个非常重要的问题，因为一个坏的选择可能会导致粒子样本的恶化，从而浪费计算能力。在本文中，我们介绍了一种新的算法自适应逼近所谓的最佳建议核的混合物的综合曲线指数分布与逻辑权重。这一系列的分布，被称为asmixtures的专家，是广泛的，足以用于在存在多模态或强烈偏态分布。的混合物，通过在线EM方法，通过最小化的辅助目标和仪器的粒子滤波器的分布之间的Kullback-Leibler分歧的最佳内核。在粒子滤波器的每次迭代中，该算法只需要解决整个粒子样本的单个优化问题，从而产生仅具有线性复杂度的算法。此外，我们在仿真研究中说明了该方法可以成功地应用于非线性状态空间模型的最优滤波。
Appropriately designing the proposal kernel of particle filters is an issue of significant importance, since a bad choice may lead to deterioration of the particle sample and, consequently, waste of computational power. In this paper we introduce a novel algorithm adaptively approximating the so-called optimal proposal kernel by a mixture of integrated curved exponential distributions with logistic weights. This family of distributions, referred to asmixtures of experts, is broad enough to be used in the presence of multi-modality or strongly skewed distributions. The mixtures are fitted, via online-EM methods, to the optimal kernel through minimisation of the Kullback-Leibler divergence between the auxiliary target and instrumental distributions of the particle filter. At each iteration of the particle filter, the algorithm is required to solve only a single optimisation problem for the whole particle sample, yielding an algorithm with only linear complexity. In addition, we illustrate in a simulation study how the method can be successfully applied to optimal filtering in nonlinear state-space models.