ABSTRACT
This chapter introduces support vector machine (SVM). It is one of machine learning methods and is one of gradient based/detrended spatial predictive methods that can use predictive variables. The functions svm and tune.svm in the e1071 package are used for SVM modeling. The function svmcv in the spm2 package is used to demonstrate the parameter estimation and accuracy assessment of SVM predictive models. And the predict function is introduced for generating the spatial predictions of SVM. The hard data set in the spm package is used to show the application of SVM to categorical data. The sponge2 point data set in the spm2 package and the sponge.grid grid data set in spm are used to show the application of SVM to numerical data.
