Сравнение методов

	Динамический вариационный вывод	Фильтр частиц (последовательное Монте-Карло)
Область	Байесовские методы	Байесовские методы
Семейство	Bayesian methods	Bayesian methods
Год появления	2014–2015	1993
Автор метода	Bayer, Osendorfer, Krishnan and colleagues	Gordon, Salmond & Smith
Тип	Bayesian approximate inference	Sequential Monte Carlo estimator
Основополагающий источник	Krishnan, R. G., Shalit, U., & Sontag, D. (2015). Deep Kalman Filters. NIPS 2015 Workshop on Advances in Approximate Bayesian Inference.	Gordon, N. J., Salmond, D. J., & Smith, A. F. M. (1993). Novel approach to nonlinear/non-Gaussian Bayesian state estimation. IEE Proceedings F (Radar and Signal Processing), 140(2), 107–113.
Другие названия	sequential variational inference, temporal variational inference, variational inference for state-space models, DVI	SMC, sequential Monte Carlo, bootstrap filter, condensation algorithm
Связанные	6	4
Сводка	Dynamic variational inference extends the variational inference framework to sequential and time-series settings by positing a structured approximate posterior that respects the temporal ordering of latent states. It jointly learns a generative model of how hidden states evolve over time and a recognition network that maps observed sequences back to those latent states, optimising a sequential evidence lower bound (ELBO).	The particle filter, introduced by Gordon, Salmond, and Smith in 1993, is a sequential Monte Carlo algorithm that approximates the Bayesian filtering distribution for nonlinear and non-Gaussian state-space models. Rather than tracking a single best estimate, it maintains a cloud of N weighted random samples — particles — that collectively represent the full posterior distribution of a hidden state at each point in time as new observations arrive.