RSVM.Anal: Support Vector Machine (SVM)

In flajole/MApckg: R package of MetaboAnalyst functions

Description

Perform recursive SVM for feature selection and classification. Uses svm function. Writes results to "svm_sigfeatures.csv" file. R-SVM uses SVM (with linear kernel) to perform classification recursively using different feature subsets. Features are selected based on their relative contribution in the classification using cross validation error rates. The least important features are eliminated in the subsequent steps. This process creates a series of SVM models (levels). The features used by the best model are plotted. LOOCV: leave one out cross-validation.

Usage

RSVM.Anal(dataSet, analSet, cvType = 10)

Arguments

dataSet	List, data set object generated by MS_to_MA function.
analSet	List, containing the results of statistical analysis (can be just an empty list).
cvType	Type of cross-validation of: integer - N-fold-CV is used; "LOO" - leave-one-out CV (LOOCV); "bootstrape" - bootstrape CV

Value

Native analSet with one added element $svm containing - standard randomForest function output

See Also

svm for used statistical function
PlotRSVM for plotting functions

flajole/MApckg documentation built on May 16, 2019, 1:16 p.m.