Markov chain Monte Carlo (MCMC) estimates a Bayesian posterior through repeated sampling when exact calculation is impractical. It makes inference possible in problems with thousands or millions of unknown quantities, but the result remains dependent on the assumptions, data, and judgments behind the model. section 2.5
Markov chain monte carlo
Markov chain Monte Carlo is a family of algorithms that uses Markov chains to generate samples from probability distributions.
How Timeless Algorithms, The Foundational Ideas covers it
- How deep it goes
- Explains repeated sampling and its role in posterior estimation, but does not detail specific algorithms or sampling mechanics.
- Examples
- Names genomics, climate science, financial modeling, and engineering design as application areas.
- What you can do after
- Understand how repeated sampling can estimate a posterior when exact solutions are impractical.
- How it connects
- Connects approximate inference to Bayesian inference and posterior distributions.
Also explained in
-
Grokking BayesMarkov chain Monte Carlo is a family of sampling algorithms that constructs a Markov chain with a target probability distribution as its stationary distribution.
-
Machine Learning Algorithms in DepthMarkov chain Monte Carlo is a Monte Carlo methodology that samples a Markov chain to approximate a target probability distribution, typically a posterior distribution.
-
Causal AIMarkov chain Monte Carlo is a family of sampling algorithms that generate a Markov chain whose limiting distribution is a target probability distribution.
Compare how 4 books explain markov chain monte carlo →
Related concepts
- metropolis-hastings samplingincludes · Metropolis-Hastings sampling is a Markov chain Monte Carlo method that proposes candidate states and accepts or rejects them with probabilities chosen to preserve a target distribution.
- variational inferencealternative to · Variational inference is an approximate inference technique that optimizes a tractable family of probability distributions to approximate a target distribution, often a posterior.
- datauses · Data is recorded or represented information, observations, measurements, or values that can be stored, processed, communicated, or interpreted.
- grid approximationalternative to · A grid approximation estimates a distribution by evaluating and normalizing probability weights on a finite grid of values.
- stochastic variational inferencealternative to · Stochastic variational inference optimizes a variational approximation using stochastic estimates of an objective or its gradient, often computed from data subsets.
- unnormalized posterioruses · A probability or distribution proportional to the posterior, given by the prior times the likelihood before normalization.
This explainer was generated by Manning's knowledge graph from the text of Timeless Algorithms, The Foundational Ideas. The section links point to the parts of the book it draws on.



