Comparer des méthodes

	Inférence variationnelle pour séries temporelles	MCMC pour séries temporelles
Domaine	Bayésien	Bayésien
Famille	Bayesian methods	Bayesian methods
Année d'origine	1999–2017	1994–1997
Auteur d'origine	Jordan, Ghahramani, Jaakkola, Saul; extended by Blei and colleagues	Carter & Kohn; West & Harrison
Type	Approximate Bayesian inference	Bayesian posterior sampling for time-ordered data
Source fondatrice	Blei, D. M., Kucukelbir, A. & McAuliffe, J. D. (2017). Variational inference: A review for statisticians. Journal of the American Statistical Association, 112(518), 859-877.	Carter, C. K. & Kohn, R. (1994). On Gibbs sampling for state space models. Biometrika, 81(3), 541–553.
Alias	time-series VI, variational Bayes for time series, TSVI, sequential variational inference	MCMC time series, Bayesian time series MCMC, time series posterior sampling, sequential Bayesian MCMC
Apparentées	6	6
Résumé	Time series variational inference applies variational Bayes to sequential data, approximating the intractable posterior over latent states and parameters with a tractable family of distributions. By maximising the evidence lower bound (ELBO), it delivers fast, scalable Bayesian inference for state-space models, dynamic latent variable models, and other time-ordered probabilistic systems.	Time series MCMC applies Markov chain Monte Carlo methods to Bayesian inference over time-ordered data. Rather than optimising a single parameter estimate, it draws samples from the full joint posterior of parameters and latent states, yielding probability distributions that honestly reflect uncertainty about dynamics, trends, and seasonal patterns across every time point.