MLC++ Utilities README file. Machine Learning Library in C++. Keywords: machine learning, accuracy estimation, cross-validation, bootstrap, ID3, decision trees, decision graphs, naive-bayes, decision tables, majority, induction algorithms, classifiers, categorizers, general logic diagrams, instance-based algorithms, discretization, lazy learning, bagging. Contact: Ronny Kohavi ronnyk@sgi.com http://robotics.stanford.edu/~ronnyk Questions or help requests related to the utilities should be addressed to mlc@postofc.corp.sgi.com Please see the MLC++ home page first: http://www.sgi.com/Technology/mlc ______________________________________________________________________ Quick starter: The MLC++ utilities are accessible through our web page http://www.sgi.com/Technology/mlc cd zcat | tar xvf - (or "tar xvf -" if you got have the .tar file) where is the directory under which the mlc directory will be built (e.g., /usr/local), and is the kit appropriate for your machine. The documentation is in utils.ps. The environment variable MLCDIR must be set to the directory where the utilities are installed, and MLCPATH to the directory where the databases are stored (can be colon-separated paths). Other optional environment variables are: OPTION_DUMP, the file to dump used options into so they can be sourced, and TMPDIR for the directory to put temporary files into. Example setup file (can be sourced in .login). setenv MLCDIR "/usr/local/mlc" setenv MLCPATH ".:$MLCDIR/db" setenv OPTION_DUMP ~/.mlcoptions setenv TMPDIR "/usr/tmp" If you have not "registered" through the web page, please send the registration form mlc-register.txt to us. Databases in the MLC++ format, which is very similar to C4.5 format can be found in starry.stanford.edu/pub/ronnyk/mlc/db. Most datafiles are converted from the repository at UC Irvine. ______________________________________________________________________ Longer guide: 1. MLC++ utilities are distributed in object code and can only run on SGI. An older version exists for suns and you can recompile it on any machine with a good compiler. 2. Get the kit from our home page under utilities. 3. Change to the directory where you want to install MLC++. If you have "root" access, "cd /usr/local"; otherwise, "cd ~" 4. Extract the files by typing zcat | tar xvf - 5. Type the following: setenv MLCDIR /mlc where is the directory you're in. 6. Type the following setenv MLCPATH ".:$MLCDIR/db" to set the default search directories to current and $MLCDIR/db. Consider adding the two setenv statements to to your .login. 6. Print the documentation for the utilities by doing lpr mlc/util.ps 7. If you have not "registered" yet, please do so through url: http://www.sgi.com/Technology/mlc/mail.html 8. Remember that many databases at UC Irvine have been converted to the standard MLC++/C4.5 format and can be retrieved from starry.stanford.edu/pub/ronnyk/mlc/db by anon ftp.