This is the README for the JAVA-package of the logistic regression
"Long Short-Term Memory" for DNA classification.

:: License

LICENSE, WARRANTY, AND LIABILITY
This programm is freely available under the GNU General Public License
(GPL).
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc., 51
Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.

:: Information

LSTM logistic regression is stripped down LSTM which can be interpreted
as logistic regression with the spectrum kernel for sequence classification.

For any step in the DNA sequence and a given k a k-mer string vector is
build and fed into the network. The LSTM architecture is just two memory
cells and no input- or output gates. The memory cells are not connected
with each other. The squashing function h is the identity function.

For testing a DNA data set is included.

:: Parameters

Important parameters are set in the JLSTM.logreg.properties file. The command line
parameters are the same as in the JLSTM (see README)

:: Running

(Unix environment)

java -classpath "lib/biojava.jar:lib/bytecode.jar:jlstm.jar" at.jku.bioinf.jlstmlogregression.JLSTM -p JLSTM.logreg.properties -w wm.mat -t 4

(Windows environment)

java -classpath "lib/biojava.jar;lib/bytecode.jar;jlstm.jar" at.jku.bioinf.jlstmlogregression.JLSTM -p JLSTM.logreg.properties -w wm.mat -t 4

This starts the logistic regression jLSTM with four threads (-t 4).