Parametric Inference for imperfectly observed Gibbsian fields

By Laurent Younes. 1989

Subfields: Probabilistic & Bayesian Methods

How researchers see this paper

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

In Probabilistic & Bayesian Methods

Citations in the corpus

Neural Autoregressive Distribution Estimation (2016, methodology): “Another line of development has been to optimize likelihood using RobbinsMonro stochastic approximation (Younes, 1989), also known as Persistent CD (Tieleman, 2008), and develop good MCMC samplers for deep undirected models (Salakhutdinov, 2009, 2010; Desjardins et al., 2010; Cho et al., 2010).”
Herding Dynamic Weights for Partially Observed Random Field Models (2009, background): “Our “herding dynamics” does not require expensive operations such as exponentiation and is fully deterministic.”
Learning with Hierarchical-Deep Models (2013, background): “Note the close connection between the form of the mean-field fixed point updates and the form of the conditional distribution3 defined by (2).”
DVAE++: Discrete Variational Autoencoders with Overlapping Transformations (2018, background): “Instead of manually coding the gradient of the negative energy function for each sample and modifying the gradient of whole objective function, we compute the negative average energy on L samples (indexed by l ) generated from PCD chains ( − .”
Deep Boltzmann Machines (2009, background): “SAP belongs to the class of well-studied stochastic approximation algorithms of the Robbins–Monro type (Robbins and Monro, 1951; Younes, 1989, 2000).”
On Herding and the Perceptron Cycling Theorem (2010, background): “For example, we may start the local search from the state we ended up in during the previous iteration (a so-called persistent chain [13, 17]).”
Noise-Contrastive Estimation of Unnormalized Statistical Models, with Applications to Natural Image Statistics (2012, methodology): “…the trade-off in noise-contrastive estimation, and compare it to the trade-off in Monte Carlo maximum likelihood estimation (Geyer, 1994), contrastive divergence (Hinton, 2002) andpersistent contrastive divergence5 (Younes, 1989; Tieleman, 2008), as well as score matching (Hyvärinen, 2005).”
Inductive Principles for Restricted Boltzmann Machine Learning (2010, methodology): “Younes (1989) showed that one can instead alternate between simulating the Markov chain for a single step and performing a parameter update using a small learning rate.”
Learning Trans-Dimensional Random Fields with Applications to Language Modeling (2018, background): “Moreover, we further examine related work in language modeling, which is our motivating application in trans-dimensional modeling.”
Stochastic gradient estimate variance in contrastive divergence and persistent contrastive divergence (2013, methodology): “Popular methods to train Restricted Boltzmann Machines [1] include Contrastive Divergence [2, 3] and Persistent Contrastive Divergence1 [4, 5].”
Training Restricted Boltzmann Machines on Word Observations (2012, methodology): “This procedure belongs to the general class of Robbins-Monro stochastic approximation algorithms (Younes, 1989).”
Trans-dimensional Random Fields for Language Modeling (2015, background): “Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing, pages 785–794, Beijing, China, July 26-31, 2015. c©2015 Association for Computational Linguistics”