File: /afs/cs.wisc.edu/p/ftp/math-prog/cpo-dataset/machine-learn/checker/README
This directory contains six files:

README (this file),  checker.mat  and checker.txt, checkerboard.eps, poly6.eps and sin.eps.   

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

This dataset is made public with the permission of its originators:

@inproceedings{hk:96,
	author = "Tin Kam Ho and Eugene M. Kleinberg",
	title =  "Building Projectable Classifiers of Arbitrary Complexity",
	booktitle={Proceedings of the 13th International Conference on Pattern R
		ecognition},
	editor = "",
	publisher = "",
	note = {http://cm.bell-labs.com/who/tkh/pubs.html},
	address = {Vienna, Austria },
	pages = {880-885},
	date = {August 25--30},
	year = 1996}

This dataset has been used in the following papers:

@incollection{Kaufman98a,
	author = {L. Kaufman},
	title={Solving the Quadratic Programming problem arising in support
  	             vector classification},
	editor = {Bernhard {Sch\"olkopf} and Christopher J. C. Burges and
                                        Alexander J. Smola},
                     booktitle = {Advances in Kernel Methods {-} Support Vector Learning},
                     publisher = {MIT Press},
	pages = {147-167},
                     year= {1999} }


@techreport{mm:99,
	author = "O. L. Mangasarian and David R. Musicant",
	title =  "Data Discrimination via Nonlinear Generalized Support
	                 Vector Machines",
	institution =  uwcs,
	month = {March},
	year =         1999,
	number =       {99-03},
	address =      "Madison, Wisconsin",
	note={ftp://ftp.cs.wisc.edu/math-prog/tech-reports/99-03.ps}}


@techreport{lm:99,
	author = "Yuh-Jye Lee and O. L. Mangasarian",
	title =  "{SSVM}: A Smooth Support Vector Machine",
 	institution =  "Data Mining Institute, Computer Sciences Department, University of Wisconsin",
	month = {September},
	year =         1999,
	number =       {99-03},
	address =      "Madison, Wisconsin",
	note={ftp://ftp.cs.wisc.edu/pub/dmi/tech-reports/99-03.ps}}

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

checker.mat
----------

This file can be loaded into MATLAB directly and contains two matrices, A and B.

A is 486 x 2 matrix which represents 486 points of class I  in R^2 space.

B is 514 x 2 matrix which represents 514 points of class II in R^2 space.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

checker.txt
----------

This text file contains three columns which represent 1000 points in R^2,
the two dimensional real space.

The first column is the indicator of classes. 

The second and third column are the coordinates of the data points.   


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

checkerboard.eps
--------------

This eps file contains the figure of checkerboard dataset.


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

poly6.eps
--------

This eps file depicts the result of checkerboard problem using SSVM with sixth drgree polynomial kernel.
SSVM stands for Smooth Support Vector Machine.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

sin.eps

This eps file depicts the result of checkerboard problem using SSVM with sinusoidal kernel.