Journal of the American Statistical Association · 1999 · 839 citations · 0 references
ComputingPhysicsStatisticsStochastic ModelingStochastic ProcessesBayesian FilteringBayesian InferenceMarkov Chain Monte CarloMonte Carlo MethodsMonte Carlo SimulationParticle FilteringAuxiliary Particle FilteringFilteringParticle FiltersPrior Distribution SpecificationSequential Importance Resampling
Abstract This article analyses the recently suggested particle approach to filtering time series. We suggest that the algorithm is not robust to outliers for two reasons: The design of the simulators and the use of the discrete support to represent the sequentially updating prior distribution. Here we tackle the first of these problems. Key Words: FilteringMarkov chain Monte CarloParticle filterSampling/importance resamplingSimulationState space