SHOGUN 0.6.4 (Default branch)


 
Thread Tools Search this Thread
Special Forums News, Links, Events and Announcements Software Releases - RSS News SHOGUN 0.6.4 (Default branch)
# 1  
Old 08-15-2008
SHOGUN 0.6.4 (Default branch)

ImageSHOGUN is a machine learning toolbox whose focus is on large scale kernel methods and especially on Support Vector Machines (SVM). It provides a generic SVM object interfacing to several different SVM implementations, all making use of the same underlying, efficient kernel implementations. Apart from SVMs and regression, SHOGUN also features a number of linear methods like Linear Discriminant Analysis (LDA), Linear Programming Machine (LPM), (Kernel) Perceptrons, and algorithms to train hidden Markov models. SHOGUN can be used from within C++, Matlab, R, Octave, and Python.License: GNU General Public License (GPL)Changes:
This release contains major feature enhancements and bugfixes. It implements 2-norm Multiple Kernel Learning, has greatly extended documentation, adds a Gaussian kernel for 32-bit floating point features, and implements the test suite for most of the functions for most interfaces. It also fixes a bug in filtering out duplicate signals in the signal handler, and fixes random number generator initialization.Image

More...
Login or Register to Ask a Question

Previous Thread | Next Thread
Login or Register to Ask a Question
svm-scale(1)							   User Manuals 						      svm-scale(1)

NAME
svm-scale - scale data to a restricted range as preprocessing for SVM training SYNOPSIS
svm-scale [ -l lower ] [ -u upper ] [ -y y_upper ] [ -s save_filename ] [ -r restore_filename ] datafilename DESCRIPTION
svm-scale reads the given datafilename (a training or testing data file as specified in svm_train(1) or svm_predict(1) ) and scales all dimensions to the given ranges. OPTIONS
-l lower lower is the lowest (minimal) value allowed in each dimension. It defaults to -1. -u upper upper is the highest (maximal) value allowed in each dimension. It defaults to 1. -y y_lower y_lower is a boolean value (0 or 1) indicating whether or not y-values (targets) should be scaled. It defaults to 0. -s save_filename save_filename indicates the filename to save the scaled data to. -r restore_filename restore_filename indicates the filename reserved to hold the original (unscaled) data in case there is a need to restore. FILES
datafilename must be a training or testing dataset. ENVIRONMENT
No environment variables. DIAGNOSTICS
None documented; see Vapnik et al. BUGS
Please report bugs to the Debian BTS. AUTHOR
Chih-Chung Chang, Chih-Jen Lin <cjlin@csie.ntu.edu.tw>, Chen-Tse Tsai <ctse.tsai@gmail.com> (packaging) SEE ALSO
svm-train(1), svm-predict(1) Linux MAY 2006 svm-scale(1)