Nested Sampling

Nested sampling is a computational method for both generating samples from a posterior probability distribution and calculating the Bayesian evidence. The approach was invented by John Skilling in 2004 primarily for computing Bayesian evidence values, but nested sampling is now also widely used for parameter estimation from samples. Parameter estimation with nested sampling often performs well compared to alternative methods like Markov chain Monte Carlo for posterior distributions with multiple local maxima. Popular nested sampling software packages include MultiNest, PolyChord, and dynesty.


Applications of Bayesian statistics in research can generally be divided into model selection and parameter estimation. Model selection involves working out which model of the data is most likely to be correct. This can be done by comparing the Bayesian evidence values for each ...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles