Interacting Particle Inferencefor Probabilistic Programming in Haskell
University essay from Uppsala universitet/Institutionen för informationsteknologi
Abstract: Probabilistic programming shows much promise as a declarative way to define statistical models, but inference is often expensive. A parallelisable particle Markovchain Monte Carlo sampler is implemented in Haskell and the domain-specific language Monad-Bayes. The method shows good performance compared to a single SMC sampler, but the full potential of the method could not be acheived.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)
