OPTIMAL PERCEPTUAL INFERENCE

By Geoffrey E. Hinton, J. Sejnowski. 1983

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

Lightweight Probabilistic Deep Networks (2018, background): “Perhaps among the ﬁrst probabilistic neural networks are the works relating undirected graphical models to neural networks via Boltz-mann machines [3, 24]; similarly, links are established for deep belief nets (DBN) via sigmoid belief nets [23, 41].”
Learning with Hierarchical-Deep Models (2013, methodology): “The original learning algorithm for Boltzmann machines used randomly initialized Markov chains to approximate both expectations to estimate gradients of the likelihood function [14].”
Direct Feedback Alignment Provides Learning in Deep Neural Networks (2016, methodology): “Boltzmann machine learning in different variants are biologically inspired methods for training neural networks, see [6], [10] and [5].”
Bidirectional Helmholtz Machines (2015, background): “As a few examples of deep models, Geoffrey Hinton and Terry Sejnowski in 1985 propose Boltzmann machines (BMs) which are undirected graphical model, in which all connections between layers are undirected [38, 31, 37].”
Serial Order: A Parallel Distributed Processing Approach (1997, background): “These intermediate units will be referred to as hidden units, following the terminology of Hinton and Sejnowski (1983).”
Perturb-and-MAP random fields: Using discrete optimization to learn and sample from energy models (2011, background): “The model sufficient statistics can be computed exactly in tree-structured (and low tree-width) graphs, but in general graphs one needs to resort to MCMC techniques for approx imating them [10, 11,31], an avenue considered too costly for many computer vision applications.”
Binary models for marginal independence (2007, background): “They focused on the quadratic binary exponential distribution, which is also known as the Boltzmann machine (Hinton and Sejnowski, 1983) or the autologistic scheme (Besag, 1974).”
Computational psychiatry (2011, methodology): “Inspired by Hopfield's work and the seminal (and still classic) work of Stuart and Donald Geman on Gibbs sampling and Bayesian approaches to image analysis, Hinton and Sejnowski [34] showed that probabilistic activation in simple units could perform a sophisticated Bayesian style of inference.”
An Efficient Learning Procedure for Deep Boltzmann Machines (2012, methodology): “The parameter updates, originally derived by Hinton and Sejnowski (1983), that are needed to perform gradient ascent in the log likelihood can be obtained from equation 2.2, where α is a learning rate.”