+--------------------------+
|           CSI            |
+--------------------------+


Version 1.0 - May 26th, 2004
----------------------------------



Description
-----------
The csi package is a Matlab program that implements the predictive 
low-rank decomposition for kernel matrices. The algorithm takes into 
account side information such as class labels or response variables 
in order to build the approximation of the kernel matrix

For more information, please read the following paper:

Francis R. Bach, Michael I. Jordan. Predictive low-rank decomposition 
for kernel methods. Proceedings of the Twenty-second International 
Conference on Machine Learning (ICML), 2005.


The csi package is Copyright (c) 2005 by Francis Bach. If you have any 
questions or comments regarding this package, or if you want to report 
any bugs, please send an e-mail to francis.bach@mines.org. The current 
version 1.0 has been released on May 26th 2005. 





Installation
------------

1. Unzip all the .m files in the same directory, and add the directory to the
matlab path.


2. (Optional) if you want a faster implementation which  C code: 
at the matlab prompt, in the directory where the package is
installed, type:

 >> mex csi.c
 >> mex csi_gaussian.c

It should create compiled files whose extensions depends on the platform
you are using:
      Windows: csi.dll,  csi_gaussian.dll 
      Linux: csi.mexglx, csi_gaussian.mexglx
 
To check if the file was correcly compiled, type

 >> which csi
 >> which csi_gaussian

and the name of the compiled versions should appear. If you have any
problems with the C file of if you are using a platform not mentioned,
please e-mail francis.bach@mines.org.

Note that compiled versions for Linux and Windows are included.




How to use the CSI package
--------------------------

The main function that you should use to run the algorithm are 
'csi' and 'csi_gaussian'. A demonstration script is included
as 'demo_csi.m'.










