
----------------------------------------------------------
			ANNOUNCEMENT

	OC1 is now available via Anonymous FTP.
----------------------------------------------------------

     OC1 (Oblique Classifier 1) is a multivariate decision tree
induction system designed for applications where the instances have
numeric feature values.  OC1 builds decision trees that contain linear
combinations of one or more attributes at each internal node; these
trees then partition the space of examples with both oblique and
axis-parallel hyperplanes.  OC1 has been used for classification of
data from several real world domains, including astronomy, cancer
diagnosis, and other.  A technical decription of the algorithm can be
found in the AAAI-93 paper by Sreerama K.  Murthy, Simon Kasif, Steven
Salzberg and Richard Beigel.  A postscript version of this paper is
provided with the package.

     OC1 is a serial version written entirely in ANSI C. OC1
incorporates a number of features intended to support flexible
experimentation on real and artificial data sets.  We have provided
support for cross validation experiments, generation of artificial
data, and graphical display of data sets and decision trees.  The OC1
software allows the user to create both standard, axis-parallel decision
trees and oblique (multivariate) trees.


        TO OBTAIN OC1 BY ANONYMOUS FTP
        --------------------------------

     The latest version of OC1 is available free of charge, and may be 
obtained via anonymous FTP from the Johns Hopkins University Computer 
Science Department.

     To obtain a copy of OC1, type the following commands:

     UNIX_prompt>  ftp blaze.cs.jhu.edu
[Note: the Internet address of blaze.cs.jhu.edu is 128.220.13.50]
     Name: anonymous
     Password: [enter your email address]

     ftp>  bin
     ftp>  cd pub/oc1
     ftp>  get oc1.tar.Z
     ftp>  bye

[Place the file oc1.tar.Z in a convenient subdirectory.]

     UNIX_prompt> uncompress oc1.tar.Z
     UNIX_prompt> tar -xf oc1.tar

[Read the file "README", to get cues to other documentation files, and
 to run the programs.]


If you have any comments, questions or suggestions, please contact

       Sreerama K. Murthy or
       Steven Salzberg or
       Simon Kasif
       Department of Computer Science
       The Johns Hopkins University
       Baltimore, MD 21218
       Email:  murthy@cs.jhu.edu,salzberg@cs.jhu.edu,kasif@cs.jhu.edu

OC1 IS INTENDED FOR NON-COMMERCIAL PURPOSES ONLY.  OC1 may be used,
copied, and modified freely for this purpose.  Any commercial use of
OC1 is strictly prohibited without the express written consent of
Sreerama K. Murthy, Simon Kasif, and Steven Salzberg, at the
Department of Computer Science, Johns Hopkins University.

-----------------------------------------------------------------------
From murthy Thu Nov  4 18:52:31 1993
Date: Thu, 4 Nov 93 18:52:00 EST
To: salzberg, kasif, murthy
Subject: OC1: modifications
Cc: rupali@jhufos.pha.jhu.edu, hcp2x@honi3.acc.virginia.edu,
        ncv@tci002.uibk.ac.at, brodley@cs.umass.edu
Content-Length: 948
X-Lines: 23
Status: RO

Hi,

I have recently modified OC1 in the following ways:
  1. Replaced the ad-hoc pruning strategy with a standard pruning procedure
     (Breiman et al's Error Complexity pruning using a secondary test set).
  2. Accelerated the oblique decision tree inducing program so that it runs in
     about half the time that it used to take before.
  3. Included implementations for two more impurity measures, namely,
     gini_index and the twoing rule.

The modified version of OC1, along with the updated documentation, is available
for anonymous FTP from blaze.cs.jhu.edu, in the directory pub/oc1, as the 
compressed tar file oc1.tar.Z. Please make a note of this change, and
obtain the new version, if you plan to use OC1. Also, could you please let 
anyone, that you know has FTPed the old version of OC1, know about this
updation ?  

Thank you.

-Sreerama K. Murthy
 Department of Computer Science
 Johns Hopkins University
 Baltimore, MD 21218

