1 /** \mainpage SHOGUN Project Documentation
5 \image html shogun_logo.png
7 \section intro_sec Introduction
8 SHOGUN is a large scale machine learning toolbox with focus on especially
9 Support Vector Machines (SVM). It provides a generic SVM object interfacing
10 to several different SVM implementations, among them the state of the art
11 LibSVM, SVMLight, SVMLin and GPDT. Each of the SVMs can be combined with a
12 variety of kernels. The toolbox not only provides efficient implementations
13 of the most common kernels, like the Linear, Polynomial, Gaussian and Sigmoid
14 %Kernel but also comes with a number of recent string kernels as e.g. the
15 Locality Improved, Fischer, TOP, Spectrum, Weighted Degree %Kernel (with
16 shifts). For the latter the efficient LINADD optimizations are
17 implemented. Also SHOGUN offers the freedom of working with custom
18 pre-computed kernels. One of its key features is the combined kernel which
19 can be constructed by a weighted linear combination of a number of
20 sub-kernels, each of which not necessarily working on the same domain. An
21 optimal sub-kernel weighting can be learned using Multiple %Kernel Learning.
22 Currently SVM 2-class classification and regression problems can be dealt
23 with. However SHOGUN also implements a number of linear methods like Linear
24 Discriminant Analysis (LDA), Linear Programming Machine (LPM), (%Kernel)
25 Perceptrons and features algorithms to train hidden markov models. The input
26 feature-objects can be dense, sparse or strings and of type
27 int/short/double/char and can be converted into different feature types.
28 Chains of preprocessors (e.g. subtracting the mean) can be attached to each
29 feature object allowing for on-the-fly pre-processing.
31 SHOGUN is implemented in C++ and interfaces to Matlab(tm), R, Octave and
32 Python \ref interfaces "(see Interfaces)".
34 \li \subpage installation
35 \li \subpage screenshots
38 \li \subpage methods "Implemented Methods"
39 \li \subpage interfaces