Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper

By Arthur P. Dempster, Nan M. Laird, Donald B. Rubin. 1977

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

End-to-End Training of Multi-Document Reader and Retriever for Open-Domain Question Answering (2021, background): “Expectation-maximization (Dempster et al., 1977) offers a principled template for learning this class of latent variable models.”
Your Classifier is Secretly an Energy Based Model and You Should Treat it Like One (2019, background): “1 INTRODUCTION For decades, research on generative models has been motivated by the promise that generative models can benefit downstream problems such as semi-supervised learning, imputation of missing data, and calibration of uncertainty (e.g., Chapelle et al. (2006); Dempster et al. (1977)).”
Representation Learning: A Review and New Perspectives (2012, methodology): “In the presence of latent variables, expectation maximization (Dempster et al., 1977) is employed where the parameters are optimized with respect to the marginal likelihood, i.”
nocaps: novel object captioning at scale (2019, methodology): “These complete captions are then used as training targets in an iterative algorithm inspired by expectation maximization (EM) [7].”
Unsupervised Word Sense Disambiguation Rivaling Supervised Methods (1995, neutral): “For example: °Including variants of the EM algorithm (Bantu, 1972; Dempster et al., 1977), especially as applied in Gale, Church and Yarowsky (1994).”
Optimization Methods for Large-Scale Machine Learning (2016, methodology): “re important methods that are not included in our presentation|such as the alternating direction method of multipliers (ADMM) [57, 64,67] and the expectation-maximization (EM) method and its variants [48,160]|but our study covers many of the core algorithmic frameworks in optimization for machine le”
Stochastic variational inference (2012, methodology): “The updates also show a connection to the expectation-maximization (EM) algorithm (Dempster et al., 1977)—Equation 16 corresponds to the E step, and Equation 15 corresponds to the M step (Neal and Hinton, 1999).”
Seeing through the Human Reporting Bias: Visual Classifiers from Noisy Human-Centric Labels (2015, methodology): “Since our model includes latent variables z for each concept and image, one approach could be to use Expectation Maximization (EM) [7].”
A Systematic Comparison of Various Statistical Alignment Models (2003, methodology): “…unknown parameters θ are determined by maximizing the likelihood on the parallel training corpus: Typically, for the kinds of models we describe here, the expectation maximization (EM) algorithm (Dempster, Laird, and Rubin 1977) or some approximate EM algorithm is used to perform this maximization.”
Variational Inference: A Review for Statisticians (2016, methodology): “Neal and Hinton (1998, first published in 1993) made important connections to the expectation maximization (EM) algorithm (Dempster, Laird, and Rubin 1977), which then led to a variety of variational inference algorithms for other types of models (Waterhouse, MacKay, and Robinson 1996; MacKay 1997;…”
Towards causal generative scene models via competition of experts (2020, methodology): “A scene x is modelled with a spatial mixture model, parametrised by deep neural networks, in which learning is performed with a procedure akin to expectation maximisation (EM; Dempster et al., 1977).”
Semi-supervised learning, causality, and the conditional cluster assumption (2019, background): “While this is a hard optimization problem due to the latent variables y, a local optimum can be found via the expectation maximisation algorithm (EM) [12].”
Relational Neural Expectation Maximization: Unsupervised Discovery of Objects and their Interactions (2018, methodology): “By using Expectation Maximization (EM; Dempster et al., 1977) to compute a Maximum Likelihood Estimate (MLE) of the parameters of this mixture ( θ 1 , . . . , θ K ), we obtain a grouping (clustering) of the pixels to each object (component) and their corresponding representation.”
Continual Learning on Noisy Data Streams via Self-Purified Replay (2021, methodology): “We use the EM algorithm [15] to fit the BMM through which we obtain the posterior probability”
Nonlinear Statistical Learning with Truncated Gaussian Graphical Models (2016, methodology): “is induced by integration, expectation-maximization (EM) (Dempster et al., 1977) can be employed to facilitate inference.”
Never-Ending Learning (2015, neutral): “NELL is in an infinite loop analogous to an Expectation-Maximization (EM) algorithm 14, 30 for semi-supervised a The KB is implemented as a frame-based knowledge representation which represents language tokens (e.g., NounPhrase:bank) distinct from non-linguistic entities to which they can refer…”