uci621raw: uci621 benchmark experiment results

Description Usage Format References


Misclassification error and computation time on the data set 21 UCI Machine Learning Repository data sets of the candidate algorithms lda, rf, knn, rpart, svm and nnet.

uci621rel are precomputed relations.

Used in demo("lsbenchplot-cs621").




An array of dimension 250 x 6 x 2 x 1 (sampling x algorithms x performances x data sets).


Manuel J. A. Eugster, Torsten Hothorn and Friedrich Leisch. "Exploratory and Inferential Analysis of Benchmark Experiments", Technical Report 30, Department of Statistics, Ludwig-Maximilians-Universitaet Muenchen, Germany. 2008. http://epub.ub.uni-muenchen.de/4134/

benchmark documentation built on May 30, 2017, 5:20 a.m.