An Introduction to MCMC for Machine Learning

By Christophe Andrieu, Nando de Freitas, Arnaud Doucet, Michael I. Jordan. 2004

2,778 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Probabilistic & Bayesian Methods, Signal Processing

How researchers see this paper

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

In Artificial Intelligence

1. Sergey Levine
2. Yoshua Bengio
3. Mohit Bansal
4. Quoc V. Le
5. Pieter Abbeel
6. Percy Liang
7. Ruslan Salakhutdinov
8. Jianfeng Gao

In Probabilistic & Bayesian Methods

1. David M. Blei
2. Zoubin Ghahramani
3. Yoshua Bengio
4. Frank Hutter
5. Frank D. Wood
6. Matthew D. Hoffman
7. Richard E. Turner
8. Yura N. Perov

In Signal Processing

1. Zoubin Ghahramani
2. David M. Blei
3. Christopher R'e
4. Eric A. Wan
5. Han Liu
6. Vladlen Koltun
7. Jascha Narain Sohl-Dickstein
8. Yoshua Bengio

Citations in the corpus

• Generalizing Hamiltonian Monte Carlo with Neural Networks (2017, background): “, 1999), physics simulations (Olsson, 1995), and machine learning (Andrieu et al., 2003).”
• Neural Variational Inference for Text Processing (2015, background): “Markov chain Monte Carlo (MCMC) (Neal, 1993; Andrieu et al., 2003) and variational inference (Jordan et al.”
• Taking the Human Out of the Loop: A Review of Bayesian Optimization (2016, methodology): “MCMC methods [4] approximate the posterior with a sequence of samples that converge to the posterior; this is the approach taken in [135] on the probit model.”
• GenDICE: Generalized Offline Estimation of Stationary Values (2020, methodology): “In many areas of machine learning, Markov chain Monte Carlo (MCMC) methods are used to conduct approximate Bayesian inference by considering Markov chains whose equilibrium distribution is a desired posterior (Andrieu et al., 2002).”
• Incorporating Non-local Information into Information Extraction Systems by Gibbs Sampling (2005, methodology): “Thus it is called a Markov Chain Monte Carlo (MCMC) method; see Andrieu et al. (2003) for a good MCMC tutorial.”
• Venture: a higher-order probabilistic programming platform with programmable inference (2014, methodology): “Standard cycle or mixture hybrid kernels (see e.g. (Bonawitz, 2008; Andrieu et al., 2003; Mansinghka, 2009)) can be used to sequence Metropolis-Hastings proposals on individual variables.”
• A Tutorial on Bayesian Nonparametric Models (2011, methodology): “The idea MCMC methods is to define a Markov chain on the hidden variables that has the posterior as its equilibrium distribution (Andrieu et al., 2003).”
• A high-bias, low-variance introduction to Machine Learning for physicists (2018, neutral): “129 Input v -> h -> v’ (one pass up one pass down) 130 131 Args: 132 model: a model 133 data: a tensor of shape (num_samples, num_visible_units) 134 135 Returns: 136 tensor of shape (num_samples, num_visible_units) 137 138 """ 139 # a configuration sampled from an RBM needs to specify the values 140”
• GFlowNet Foundations (2021, neutral): “, drawing multiple chains) and simulated annealing (Andrieu et al., 2003) to facilitate jumping between modes.”
• Adaptive MCMC with Bayesian Optimization (2012, methodology): “Markov chain Monte Carlo (MCMC) algorithms are often the preferred method for accomplishing this sampling task, see e.g. Andrieu et al. (2003) and Robert & Casella (1998).”
• Guided, stochastic model-based GUI testing of Android apps (2017, methodology): “Gibbs Sampling [5, 64], is a special case of the Metropolis-Hastings algorithm [65].”