Stochastic Variational Inference for Bayesian Time Series Models

By Matthew J. Johnson, Alan S. Willsky. 2014

87 citations recorded by the source corpus.

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

Variational Inference: A Review for Statisticians (2016, background): “…2011; Neville, Ormerod, and Wand 2014), general time-series models (Roberts et al. 2004; Barber and Chiappa 2006; Archambeau et al. 2007a, 2007b; Johnson and Willsky 2014; Foti et al. 2014), robust models (Tipping and Lawrence 2005; Wang and Blei 2015), and Gaussian process models (Titsias and…”
Edward: A library for probabilistic modeling, inference, and criticism (2016, neutral): “For criticism , Edward provides methods from scoring rules (Winkler, 1996) and predictive”
Coordinated Multi-Agent Imitation Learning (2017, methodology): “This explains our optimal derivation for q ˚ p z q : q ˚ p z q9 exp q r log p 0 p z 1 qs ` t “ 2 E q p P q r log P z t ´ 1 ,z t s ` t “ 1 E q p φ q log r p p x t | z t qs Following the notation from (Johnson & Willsky, 2014), we denote p p z 1: T | x 1: T ,P,φ q “ HMM p p 0 , P, L q .”
Advances in Variational Inference (2017, methodology): “[45], [76] derive SVI for popular time series models including HMMs, hidden semi-Markov models (HSMM), and hierarchical Dirichlet process-HMMs.”
Stochastic variational inference for hidden Markov models (2014, unknown): “SVI for Bayesian time series including HMMs was recently considered in settings where each minibatch is a set of independent series [13], though in this setting again dependencies do not need to be broken.”
Structured VAEs: Composing Probabilistic Graphical Models and Variational Autoencoders (2016, methodology): “Though we do not discuss the Bayesian nonparametric case further, the algorithms we develop here immediately extend to the HDP-HMM using the methods in Johnson & Willsky (2014).”
Deep Probabilistic Programming (2017, background): “The approach extends naturally to streaming data (Doucet et al., 2000; Broderick et al., 2013; McInerney et al., 2015), dynamic batch sizes, and data structures in which working on a subgraph does not immediately apply (Binder et al., 1997; Johnson & Willsky, 2014; Foti et al., 2014).”