Theory and Use of the EM Algorithm

By Maya R. Gupta, Yihua Chen. Foundations and Trends® in Signal Processing, 2011

387 citations recorded by the source corpus.

Subfields: Control and Systems Engineering

View Paper: https://www.semanticscholar.org/paper/b2d80dde51c121c996ad7fbe2de7a7bce87dec54

How researchers see this paper

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

In Control and Systems Engineering

Citations in the corpus

An Overview of Voice Conversion and Its Challenges: From Statistical Modeling to Deep Learning (2020, methodology): “To estimate the model parameters of the JD-GMM, expectation-maximization (EM) algorithm [118]–[121] is used to maximize likelihood on the training data.”
Recommender Systems (2012, methodology): “This can be done by applying the inference methods including the marginal estimation by belief propagation [185,186] and likelihood maximization by expectation maximization [187].”
Sensitivity and Specificity of Information Criteria (2018, neutral): “To fit an LCA model or any of 349 its cousins, an algorithm such as EM (Dempster et al., 1977; McLachlan and Peel, 350 2000; Gupta and Chen, 2010) is often used to alternatively estimate class-specific 351 parameters and predict subjects’ class membership given those parameters.”
Comprehensive and reliable crowd assessment algorithms (2014, background): “B) We present experimental results on synthetic data (Section IV-B) and 3 real datasets (Section IV-C) to study the performance of our confidence intervals on k-ary tasks.”
Application-Oriented License Plate Recognition (2013, methodology): “Given the edge map [xi], the EM algorithm can be applied to obtain μj and Vj for each Gaussian cluster [16].”
Introduction to the Dirichlet Distribution and Related Processes (2010, methodology): “See [3] for details on how to use the expectation-maximization (EM) algorithm for this problem and [10] for a broader discussion including using Newton-Raphson.”
An Integrated Pipeline for de Novo Assembly of Microbial Genomes (2012, methodology): “To avoid including noise in mean and variance estimates from shadow libraries and other error sources, we perform a round of Expectation-Maximization (EM) clustering of insert sizes before calculating sample statistics [34].”
Interspecific differences in wood decay rates: insights from a new short‐term method to study long‐term wood decomposition (2012, methodology): “We generate maximum likelihood estimates for h using an algorithm conceptually similar to the expectation maximization (EM) algorithm (Dempster, Laird & Rubin 1977; Gupta & Chen 2011), that is, an iterative procedure (Fig.”