Hi: I am a newer in using R for data mining, and find the "e1071" pakage an excellent tool in doing data mining work! what frustrated me recently is that when I using the function "svm" and using the "cross=10" parameters, I got all the "accuracies" of the model greater than 1. Isn't that the accuracy should be smaller than 1? so I wander how, the accuracy of the cross validation is calculated, and what's the meaning of the accuracy? summary of the trained svm model as follows:> summary(model)Call: svm(formula = y ~ x, cachesize = 100, gamma = obj$best.parameters$gamma, cost = obj$best.parameters$cost, cross = 10) Parameters: SVM-Type: C-classification SVM-Kernel: radial cost: 4 gamma: 0.5 Number of Support Vectors: 1784 ( 892 892 ) Number of Classes: 2 Levels: high low 10-fold cross-validation on training data: Total Accuracy: 6.726457 Single Accuracies: 2.247191 7.303371 8.379888 6.179775 4.469274 6.741573 5.617978 11.73184 8.426966 6.145251 Help me urgent, waiting for your reply! Best Wishes! 2007-10-27 Sincerely! Nan Qiao Graduate student, Jing-Dong Jackie Han's Lab Institute of Genetics and Developmental Biology Chinese Academy of Sciences Datun Road, Beijing 100101 P.R. China TEL: +86-(0)10-6484 5797 [[alternative HTML version deleted]]
David Meyer
2007-Oct-28 23:07 UTC
[R] problems in cross validation of SVM in pakage "e1071"
Dear Nan Qiao, the accuracies are in percent (so times 100). Best David ------------------------ Hi: I am a newer in using R for data mining, and find the "e1071" pakage an excellent tool in doing data mining work! what frustrated me recently is that when I using the function "svm" and using the "cross=10" parameters, I got all the "accuracies" of the model greater than 1. Isn't that the accuracy should be smaller than 1? so I wander how, the accuracy of the cross validation is calculated, and what's the meaning of the accuracy?