BSD-3-Clause licensed by Manuel Bärenz
Maintained by [email protected]
This version can be pinned in stack with:rhine-bayes-1.5@sha256:e129a5797df3893ceaa4d97dc4722794377da551fb36697a7bcbb91d3a1ef01b,1939

Module documentation for 1.5

README

This package connects rhine to the monad-bayes library for probabilistic programming and inference. It provides:

  • Some standard stochastic processes such as Brownian Motion and Levý processes
  • A particle filter inference method called Sequential Monte Carlo

This allows you to do interactive probabilistic (i.e. involving randomness) programs, and at the same time perform online inference, or realtime machine learning.

An example for this is given in rhine-bayes/app/Main.hs, where inference is performed both on simulated values as well as external input given by the user.

To read more, have a look at https://www.tweag.io/blog/2023-10-12-rhine-bayes/.

Changes

Revision history for rhine-gloss

1.3

  • Dropped dunai dependency in favour of state automata. See the versions readme for details.
  • Support GHC 9.6 and 9.8
  • Updated to monad-bayes 1.3

1.2

  • Updated to monad-bayes 1.2.0

1.1

  • dunai-0.11 compatibility
  • Extended example application by particle collapse

1.0

0.9

  • Add simple Poisson, Gamma and Bernoulli processes

0.8.1.1

  • First version. Version numbers follow rhine.
  • Introduces basic stochastic processes and Sequential Monte Carlo particle filter
  • Thank you, Reuben Cohn-Gordon and Dominic Steinitz