人工智能常用工具箱【SVM，高斯过程等】

Gaussian Processes

• GP Demo. Demonstration Software for Gaussian Processes by David MacKay (in OCTAVE).
• gpml. Matlab implementations of algorithms from Rasmussen & Williams "Gaussian #此前在首页部分显示#Processes for Machine Learning", the MIT Press 2006.
• LS-SVMlab. Matlab/C toolbox for least squares support vector machines.
• MAP-1. Package for MAP estimation by Carl Rasmussen.
• MC-1. Package for MAP estimation by Carl Rasmussen.
• Flexible Bayesian Modelling. Package by Radford Neal. It includes programs for Neural Networks, Gaussian Processes, and Mixture Models.
• Netlab. Matlab toolbox including Gaussian Process Regression, Mixture models and Neural Networks.
• Sparse Gaussian Processes. Matlab Toolbox for Sparse Inference using Gaussian Processes.
• Tpros and Cpros. Package by Mark Gibbs.

Mathematical Programming

• CPLEX. Barrier/QP Solver.
• LOQO. Linear and Quadratic Optimization Package by Robert Vanderbei.
• MINOS. Linear and Quadratic Solver.
• Spire.XLS. Spire.XLS for Visual Studio .NET Free Download.

Support Vectors

• Nearest Point Algorithm. by Sathiya Keerthi (in FORTRAN).
• SVM Java Applet. by Chris Burges et al.
• BSVM. A decomposition method for bound-constrained SVM formulations.
• QP SVM Classification and Regression. Fortran Implementation.
• CLISP/LibSVM. A module for using LibSVM from GNU CLISP (an ANSI Common Lisp implementation).
• Chunking Code. by C. Saunders, M. O. Stitson, J. Weston, L. Bottou, B. Schölkopf, and A. Smola at Royal Holloway, AT&T, and GMD FIRST (Documentation).
• cSVM. SVM for classification tasks with model selection.
• 2D SVM Interactive Demo. runs under Matlab 6 and produces nice pictures - useful for courses.
• DTREG. by Phillip H. Sherrod.
• Excel2SVM. Convert Excel tables to proper format and perform SVM analysis.
• Interior Point Optimizer for SVM Pattern Recognition. by Alex Smola.
• Equbits Foresight. Commerical SVM based Classification and Regression Application Designed for Drug Discovery.
• Gini-SVM. A multi-class Probabilistic regression software for large data sets.
• GiniSVM. Multi-class SVM Probability regression package.
• Gist. Gist contains software tools for support vector machine classification and for kernel principal components analysis. The SVM portion of Gist is available via an interactive web server.
• Parallel GPDT. Parallel and serial training of SVM.
• HeroSvm1.0. A high performance DLL for training SVM on a very large training set efficiently.
• SVM java implementation. This implementation is simple and easy to modify.
• kernlab. Kernel-based Machine Learning package for R.
• LEARNSC. SVM, NN and FL MATLAB based user-friendly routines.
• LIBSVM. An SVM library with a graphic interface.
• looms. a leave-one-out model selection software based on BSVM.
• LS-SVMlab. Matlab/C Toolbox for Least Squares Support Vector Machines.
• M-SVM. Multi-class support vector machine for very large problems.
• M-SVM. Multi-class support vector machine for very large problems.
• mySVM. SVM implementation for pattern recognition and regression.
• mySVM and SVMlight for Windows. SVM implementation for Windows, uses Microsoft Visual C++ 6.0.
• mySVM/db. SVM implementation to be run inside a database.
• Sequential Minimal Optimization. by Xianping Ge.
• Online Support Vector Regression. Matlab & C++ Implementation of the Online SVR algorithm.
• Shogun. A Large Scale Machine Learning Toolbox with Focus on Kernel Methods.
• SMOBR. SMOBR is an implementation of the original Sequential Minimal Optimisation proposed by Platt written in C++.
• SVM-align. SVM training of sequence alignment models.
• SVM-CFG. SVM training for parsing with context-free grammars.
• SVM-HMM. SVM training of linear chain Hidden Markov Models.
• SVM-multiclass. Multi-class SVM.
• SVM-perf. Training linear SVMs for non-standard performance measures (e.g. F1, ROC-Area).
• SVM-QP. Convext QP solver for large-scale support vector machines classification.
• SVM-struct. SVM training for structured outputs (e.g. HMMs, grammars, sequence alignment, rankings).
• SVMdark. A Windows implementation of a support vector machine.
• SvmFu. by Ryan Rifkin.
• SVMLight. by Thorsten Joachims.
• SVM/LOO. Matlab code for SVM incremental learning and decremental unlearning (LOO validation).
• SVM/optima. SVM QP routines in Fortran for classification/regression.
• SVMseq. An implementation of grad. desc. for SV learning, supports sample selection, string kernels and quasi-linear training. Implemented in Haskell, source + binaries available.
• SVMTorch. Support Vector Machine for Large-Scale Regression and Classification Problems.
• Tiberius. A Windows based implementation of cSVM.
• Matlab SVM Toolbox. by Steve Gunn.
• Matlab SVM Toolbox. Matlab implementation in the style of SVMlight, can train 1-norm and 2-norm SVMs.
• OSU SVM Classifier Matlab Toolbox. A matlab toolbox with a C++ mex core to fast implement the SVM classifiers.
• SimpleSVM Toolbox. Fully Matlab toolbox for SVM, based on SimpleSVM algorithm. Includes 1class, invariance treatment.
• SVM Toolbox. Object Oriented MATLAB Support Vector Machine Toolbox, including C++ MEX implementation of the sequential minimal optimisation algorithm.
• WinSVM. SVM program for running under Windows.It uses SMO algorithm, so it is very fast and easy to use.
• WinSVM. SVM for windows,easy to use.
• winSVM. A Windows implementation of a support vector machine.

Other Algorithms

• AdaBoost-Reg. A regularized version of the AdaBoost algorithm (in MATLAB).
• Generalized Discriminant Analysis. Zip file, for Matlab 5.
• Kernel Billiard. by Pal Rujan (in C).
• Live Chat. Live Chat for your website. Installs in 5 minutes. Chat live with your visitors!
• Kernel ICA. A kernel-based approach for independent component analysis.
• JINFIL java Instance Filtering. Instance Filtering is a preprocessing step for supervised learning systems for entity recognition in texts. The goal of Instance Filtering is to reduce both the skewed class distribution and the data set size by eliminating negative instances, while preserving positive ones as much as possible. This process is performed on both the training and test set, with the effect of reducing the learning and classification time, while maintaining or improving the prediction accuracy. The tool demonstrate excellent performances when applied to SVM classifiers.
• Kernel-Machine Library. A GPL'ed C++ library to develop (new) kernel machine tools and algorithms in an efficient way.
• myKLR. Kernel Logistic Regression.
• RBF Networks. Fast RBF Networks with adaptive centers.
• Kernel PCA. RBF Toy Example by Bernhard Schölkopf (in MATLAB).
• R-KDDA. A regularized kernel discriminant analysis method (in matlab).
• Spider. A library in MATLAB for classification, regression, clustering, .... for SVMs it uses LIBSVM and SVMLight.
• UKR Matlab toolbox. Software package for "Unsupervised Kernel Regression", a method for learning principal manifolds. Also includes a C library for low-level functions.
• Torch. a new machine learning library in C++/GPL including MLP, RBF, SVM, GMM, HMM, KNN, Parzen...

from：http://www.kernel-machines.org/