用r做支持向量机.pptxVIP

Software reviewMost of the existing SVM software is written in C or C++ e.glibsvm SVMlight SVMTorchmySVMM-SVM第1页/共20页第一页，共21页。 Putting SVM specific software aside, there is little software available implementing a wider range of kernel methods with some exceptions like the Spider software which provides a MATLAB interface to various C/C++ SVM libraries and MATLAB implementations of various kernel-based algorithms, Torch which also includes more traditional machine learning algorithms, and the occasional MATLAB or C program found on a personal web page where an author includes code from a published paper.第2页/共20页第二页，共21页。 R packagese1071 offers an interface to libsvm provides a robust and fast SVM implementation and produces state of the art results on most classification and regression .第3页/共20页第三页，共21页。 klaR provides an interface to SVMlight a popular SVM implementation along with other classification tools like Regularized Discriminant Analysis. 第4页/共20页第四页，共21页。 LiblineaRLinear ,fasterit can be able to handle large-scaled dataset.However, most of the libsvm and klaR SVM code is in C++. Therefore, if one would like to extend or enhance the code with e.g. new kernels or different optimizers, one would have to modify the core C++ code. 第5页/共20页第五页，共21页。 Kernlabkernlab aims to allow the user to switch between kernels on an existing algorithm and even create and use own kernel functions for the kernel methods provided in the package. support vector machines the relevance vector machineGaussian processes a ranking algorithmkernel PCA kernel CCA kernel feature analysisonline kernel methods a spectral clustering algorithm. 第6页/共20页第六页，共21页。 e1071C- and ν-classificationone-class-classification (novelty detection) ε- and ν-regression linear, polynomial, radial basis function, and sigmoidal kernels K-fold Cross ValidationMulti-class classification: one-against-one第7页/共20页第七页，共21页。 e1071Svm letter.Rpredict.svmtune第8页/共20页第八页，共21页。 e1071ICABagged