Can Gibbs Sampling be used to generate synthetic data from the posterior distribution?

by boski   Last Updated July 12, 2019 09:19 AM - source

I am working with a few datasets. I would like to expand some of them somehow by creating synthetic data. Is it possible to use a Gibbs Sampler to, by sampling a given distribution of the original data, create synthetic data points?

Tags : bayesian gibbs


Related Questions


throwing away all Gibbs samples after approximation

Updated January 12, 2017 08:08 AM


Can Bayesian Optimization solve this problem?

Updated October 11, 2018 19:19 PM

Full conditionals - Gibbs Sampler

Updated March 26, 2015 07:28 AM

Gibbs Sampler - How many markov chains?

Updated March 27, 2015 13:06 PM