This classifier uses the e1071 package to implement a support vector machine.

cl_svm(ndr_container_or_object = NULL, return_decision_values = TRUE, ...)



The purpose of this argument is to make the constructor of the cl_svm classifier works with the magrittr pipe (|>) operator. This argument should almost never be directly set by the user to anything other than NULL. If this is set to the default value of NULL, then the constructor will return a cl_svm object. If this is set to an ndr container, then a cl_svm object will be added to the container and the container will be returned. If this argument is set to another ndr object, then both that ndr object as well as a new cl_svm object will be added to a new container and the container will be returned.


A Boolean specifying whether the prediction function should return columns that have the decision values. Setting this to FALSE will save memory so can be useful when analyzing very large high temporal resolution data sets. However if this is set to FALSE< metrics won't be able to compute decoding accuracy measures that are based on the decision values; e.g., the rm_main_results object won't be able to calculate normalized rank decision values.


All parameters that are available in the e1071 package svm() object should work with this CL object.


This constructor creates an NDR classifier object with the class cl_svm. Like all NDR classifier objects, this classifier will be used by a cross-validator to learn the relationship between neural activity and experimental conditions on a training set of data, and then it will be used to make predictions on a test set of data.


A support vector machine (SVM) is a classifier that learns a function f that minimizes the hinge loss between predictions made on the training data, while also applying a penalty for more complex f (the penalty is based on the norm of f in a reproducing kernel Hilbert space). The SVM has a parameter C that controls the trade off between the empirical loss (i.e., a smaller prediction error on the training set), and the complexity of the f. SVMs can use different kernels to create nonlinear decision boundaries.

SVMs are work on binary classification problems, so to do multi-class classification, an all-pairs classification scheme (which is the default for the e1071 package). In the all-pairs scheme,training separate classifiers for all pairs of labels (i.e., if there are 100 different classes then nchoosek(100, 2) = 4950 different classifiers are trained). Testing the classifier in all-pairs involves having all classifiers classify the test point, and then the class label is given to the class the was chosen most often by the binary classifiers (in the case of a tie in the number of classes that won a contest the class label is randomly chosen). The decision values for all-pairs are the number of contests won by each class (for each test point).

See also


Other classifier: cl_max_correlation(), cl_poisson_naive_bayes()


# using the default e1071 parameters
cl <- cl_svm()

# using a linear kernel
cl <- cl_svm(kernel = "linear")