Markov chain Monte Carlo (MCMC)
Markov chain Monte Carlo (MCMC) techniques are a set of statistical algorithms which aim to return an estimate of some posterior distribution in an efficient manner. In many cases, directly calculating (through "brute force methods") some probability distribution function (PDF) is computationally prohibitive and for most cases, MCMC techniques are much more computationally tractable.
The general property of MCMC techniques (what makes them "Markovian") is their lack of memory; that is, each subsequent sampling step depends only on the previous step. The sampling techniques used can
Please contact me if you would like to use this or other images or information on this website!