Software for Flexible Bayesian Modeling and Markov Chain Sampling

This software supports flexible Bayesian learning of regression, classification, density, and other models, based on multilayer perceptron neural networks, Gaussian processes, finite and countably infinite mixtures, and Dirichlet diffusion trees, as well as facilities for inferring sources of atmospheric contamination and for molecular simulation. These are implemented using Markov chain Monte Carlo methods. Facilities for Markov chain sampling from distributions specified by simple formulas for the density or for the prior and likelihood are also included.

Before trying to use the software, you may need to read various references that describe the models and the Markov chain methods used.

The software is written in C for Unix, Linux, and macOS systems.

Some of the programs in this package are designed to work wellwith my version of the 'graph' program, see github.com/radfordneal/plotutils, though the data can instead be plotted by whatever plot program you have available.

The source code is available at gitlab.

Current release of 2022年04月21日:

This release has major new features and performance improvements for neural network models, including support for convolutional networks and for computation on GPUs.

The tar archives include the complete hypertext documentation, so that you can set it up to be read from your local machine. The documentation is also readable as simple text files. For directions on how to untar a tar archive, see the installation instructions within the above hypertext documentation.

Old release of 2020年01月24日:

Old release of 2004年11月10日:

Old release of 2003年06月29日:

Old release of 2001年08月31日:

Warning: This version contains a bug in net-pred that affects predictions for multi-way classification models. It can be patched by changing line 392 of net/net-model.c, replacing "type=1" with "type==1". There's also a bug affecting the "n" option for gp-pred for multi-way classification models, which you can patch by replacing "+=" with "=" on line 440 of gp-pred.c.

Old release of 2000年08月21日:

Old release of 1999年12月06日:

Old release of 1999年03月13日:

Old release of 1998年09月01日:

Old release of 1998年08月02日:

This was the first release with the facilities for sampling from distributions specified by formulas for the "energy" or the prior and likelihood.

Old release of 1997年07月22日:

This is the release that was used for the tests in my paper on Assessing relevance determination methods using DELVE.

Old release of 1997年01月18日:

Old release of 1996年08月26日:

This is the release that was used by Carl Rasmussen for the neural network models evaluated in his thesis. Gaussian processes were not implemented in this release.

Old release of 1995年08月09日:

Back to Radford Neal's home page