This code is an example implementation of the method described in
Hierarchical Beta Processes and the Indian Buffet Process, R. Thibaux,
M. I. Jordan. Eleventh International Conference on Artificial Intelligence
and Statistics (AISTATS 2007), San Juan, Puerto Rico, 2007.

For the latest version of the code and more information, see
http://www.cs.berkeley.edu/~thibaux/Projects/AISTATS_07/intro.htm

==========================================================================
 Copyright (c) 2007, Regents of the University of California
 Created by Romain Thibaux

 All rights reserved.

 Redistribution and use in source and binary forms, with or without
 modification, are permitted provided that the following conditions are
 met:

    * Redistributions of source code must retain the above copyright
      notice, this list of conditions and the following disclaimer.
    * Redistributions in binary form must reproduce the above copyright
      notice, this list of conditions and the following disclaimer in the
      documentation and/or other materials provided with the distribution.
    * Neither the name of the University of California, Berkeley nor the
      names of its contributors may be used to endorse or promote products
      derived from this software without specific prior written
      permission.

 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
 PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER
 OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
 EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
 PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
 PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
 LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
 NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
==========================================================================


1) DATASET

The data is the 20 newsgroup dataset, available at
http://people.csail.mit.edu/jrennie/20Newsgroups/
In particular, we used the "by date" dataset at
http://people.csail.mit.edu/jrennie/20Newsgroups/20news-bydate.tar.gz

We process this data to transform each message into a binary vector, each
component of which represents the presence of a feature. The processed
data is available as a matlab file
  20news-bydate.mat
Or you can recreate the binary features by placing the files
  Makefile
  text_to_word_list.pl
in the Data/20news-bydate/ directory and running make (Windows users can
use cygwin). The script takes a while to run. The matlab function
  load_newsgroups()
can read either the .mat file or the binary vector files generated by the
script. 


2) MAIN FUNCTIONS

hbp() is the implementation of the method described in the paper.
test_hbp() compares the performance of hbp() to naive_bayes(), and in doing
so gives examples of calls to hbp().