Sequential Monte Carlo methods for dynamic systems

By Jun S. Liu, Rong Chen. 1997

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Bias Correction of Learned Generative Models using Likelihood-Free Importance Weighting (2019, methodology): “While exact sampling from p✓, is intractable, we can instead perform sample from a particle-based approximation to p✓, via sampling-importance-resampling [25, 26] (SIR).”
Monte Carlo State-Space Likelihoods by Weighted Posterior Kernel Density Estimation (2004, methodology): “Liu and West (2001) addressed the sample degradation problem using West’s (1993) approach of kernel smoothing the parameter dimensions (see also Liu and Chen 1998; and Berzuini, Best, Gilks, and Larizza 1997).”
Smoothing algorithms for state–space models (2010, methodology): “The ESS provides a measure of the degeneracy of the particle set and a larger ESS value is desired; see Liu and Chen (1998) for example.”
Sigma-point kalman filters for probabilistic inference in dynamic state-space models (2004, methodology): “…hybrid Gaussian methods (score function EKF [26], Gaussian sum filters [1]), direct and adaptive numerical integration (grid-based filters [33]), sequential Monte Carlo methods (particle filters [6] [24] [7]) and variational methods (Bayesian mixture of factor analyzers [11]) to name but a few.”
Monte Carlo filtering for multi target tracking and data association (2005, methodology): “Sequential Monte Carlo (SMC) methods [10]–[12], otherwise known as Particle Filters [8], [9], or CONDENSATION [13], have gained a lot of popularity in recent years as a numerical approximation strategy to compute the tracking recursion for complex models.”
Negative association, ordering and convergence of resampling methods (2017, methodology): “For similar reasons, the conditions of Theorem 1 are also satisfied by the residual resampling scheme of Liu and Chen (1998).”
Continual Learning with Bayesian Neural Networks for Non-Stationary Data (2020, methodology): “Prominent examples include sequential Monte Carlo (Liu & Chen, 1998), assumed density ﬁltering (Maybeck, 1982), and online variational Bayes (Opper, 1998; Ghahramani, 2000; Sato, 2001; Broderick et al., 2013).”
A Probabilistic Exclusion Principle for Tracking Multiple Objects (2000, unknown): “Evaluating the performance of particle filters such as Condensation is a difficult problem [4, 6, 12].”
An Improved Exponential Model for Predicting Remaining Useful Life of Rolling Element Bearings (2015, background): “A suitable measure of the degeneracy of particles is the effective sample size introduced in [32] and defined as follows:”
KalmanNet: Neural Network Aided Kalman Filtering for Partially Known Dynamics (2021, background): “(MC) sampling, such as the family of particle filters (PFs) [11]– [13], were introduced for state estimation in non-linear, nonGaussian SS models.”
Capturing natural hand articulation (2001, methodology): “Sequential Monte Carlo methods for dynamic systems are studied in the area of statistics [8, 9].”
Bayesian methods for neural networks (2000, methodology): “4 When to resample It has been argued in (Liu and Chen, 1995; Liu and Chen, 1998) that when all the importance weights are nearly equal, resampling only reduces the number of distinctive streams and introduces extra variation in the simulations.”
Advancing data assimilation in operational hydrologic forecasting: progresses, challenges, and emerging opportunities (2012, background): “Synthetic studies in other fields (e.g., Liu and Chen, 1998; Fearnhead and Clifford, 2003; Snyder et al., 2008) showed that PF often needs more particles than other filtering methods and the required ensemble size can increase exponentially with the number of state variables.”