Stochastic Structured Variational Inference

By Matthew D. Hoffman, David M. Blei. 2014

88 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

Bayesian Structure Adaptation for Continual Learning (2019, methodology): “We employ structured mean-ﬁeld approximation (Hoffman & Blei, 2015), which performs better than normally used mean-ﬁeld approximation, as the former captures the dependencies in the approximate posterior distributions of B and ν .”
Hierarchical Variational Models (2015, background): “Then ] In the setting where the prior has discrete variables, optimization requires a little more work. where the second to last equality follows by the derivative of the CDF function (Hoffman and Blei, 2015).”
Self-Consistent Trajectory Autoencoder: Hierarchical Reinforcement Learning with Trajectory Embeddings (2018, methodology): “We refer the reader to Hoffman & Blei (2015); Kingma & Welling (2013) for details.”
Gaussian variational approximation with sparse precision matrices (2016, background): “Structured stochastic variational inference (Hoffman and Blei, 2015) is a generalization of stochastic variational inference to allow dependencies between global and local variables.”
Semi-Implicit Variational Inference (2018, methodology): “…exists a variety of methods that add some structure back to Q to partially restore the dependencies (Saul & Jordan, 1996; Jaakkola & Jordan, 1998; Hoffman & Blei, 2015; Giordano et al., 2015; Tran et al., 2015; 2016; Han et al., 2016; Ranganath et al., 2016; Maaløe et al., 2016; Gregor et al.,…”
On Last-Layer Algorithms for Classification: Decoupling Representation from Uncertainty Estimation (2020, methodology): “Third, to consider methods for variational inference and Laplace approximations, e.g. [RBB18], and alternative methods for Bayesian logistic regression on large datasets since the focus is on last layer Bayesian approaches using the features learned from deep neural networks, e.g. [HCB16, HB15].”
On Nesting Monte Carlo Estimators (2017, background): “Hence, our results have important implications for relaxing mean-ﬁeld assumptions, or using different bounds, in variational inference (Hoffman & Blei, 2015; Naesseth et al., 2017; Maddison et al., 2017) and deep generative models (Burda et al., 2015; Le et al., 2018).”
Variational Russian Roulette for Deep Bayesian Nonparametrics (2019, methodology): “Singh et al. (2017) present a structured variational inference method for this model, based on the method of Hoffman & Blei (2015), which performs better than the more common mean-field approximation, as it introduces dependencies in the approximate posterior distribution, between Z and ν.”
Monte Carlo Gradient Estimation in Machine Learning (2019, methodology): “Hoffman and Blei (2015) used the implicit gradients to perform backpropagation through the Gamma distribution using a finite difference approximation of the CDF derivative.”