Practical filtering with sequential parameter learning

• 成果类型: Article
• 署名作者: Polson, Nicholas G.; Stroud, Jonathan R.; Mueller, Peter
• 署名单位: University of Chicago; University of Pennsylvania; University of Texas System; UTMD Anderson Cancer Center; University of Texas Health Science Center Houston
• 刊物名称: JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY
• ISSN/ISSBN: 1369-7412
• DOI: 10.1111/j.1467-9868.2007.00642.x
• 发表日期: 2008
• 页码: 413-428
• 关键词: chain monte-carlo; state-space models; likelihood inference

摘要

The paper develops a simulation-based approach to sequential parameter learning and filtering in general state space models. Our approach is based on approximating the target posterior by a mixture of fixed lag smoothing distributions. Parameter inference exploits a sufficient statistic structure and the methodology can be easily implemented by modifying state space smoothing algorithms. We avoid reweighting particles and hence sample degeneracy problems that plague particle filters that use sequential importance sampling. The method is illustrated by using two examples: a benchmark auto-regressive model with observation error and a high dimensional dynamic spatiotemporal model. We show that the method provides accurate inference in the presence of outliers, model misspecification and high dimensionality.