New software release / Dirichlet diffusion trees

[Radford Neal]

                       Announcing a new release of my 

                  SOFTWARE FOR FLEXIBLE BAYESIAN MODELING

Features include:

   * Regression and classification models based on neural networks and 
     Gaussian processes

   * Density modeling and clustering methods based on finite and infinite 
     (Dirichlet process) mixtures and on Dirichlet diffusion trees

   * Inference for a variety of simple Bayesian models specified using
     BUGS-like formulas

   * A variety of Markov chain Monte Carlo methods, for use with the 
     above models, and for evaluation of MCMC methodologies

Dirichlet diffusion tree models are a new feature in this release.
These models utilize a new family of prior distributions over
distributions that is more flexible and realistic than Dirichlet
process, Dirichlet process mixture, and Polya tree priors.  These
models are suitable for general density modeling tasks, and also
provide a Bayesian method for hierarchical clustering.  See the
following references:

   Neal, R. M. (2003) "Density modeling and clustering using Dirichlet
     diffusion trees", to appear in Bayesian Statistics 7.

   Neal, R. M. (2001) "Defining priors for distributions using Dirichlet 
     diffusion trees", Technical Report No. 0104, Dept. of Statistics, 
     University of Toronto, 25 pages.  Available at 

         http://www.cs.utoronto.ca/~radford/dft-paper1.abstract.html

The software is written in C for Unix and Linux systems.  It is free,
and may be downloaded from

         http://www.cs.utoronto.ca/~radford/fbm.software.html

----------------------------------------------------------------------------
Radford M. Neal                                       radford at cs.utoronto.ca
Dept. of Statistics and Dept. of Computer Science radford at utstat.utoronto.ca
University of Toronto                     http://www.cs.utoronto.ca/~radford
----------------------------------------------------------------------------