Benchmark evaluation of classification methods for single label learning with R

Classification in data mining is a procedure in which individual items are placed into groups based on quantitative information on one or more characteristics items and based on a training set of previously labeled items. The objective of this paper is to introduce, explain and compare the performance of the single - labeled supervised learning algorithms in R language on benchmark single labeled data set. The traditional classification algorithms like Decision Tree, Naïve Bayes, Support Vector Machine, Random Forest, Classification and Regression Trees are used under inspection. The R language is chosen to see the classification performances. Four measures (sensitivity, specificity, accuracy, F - measure) of performance here considered are based on confusion matrix, table of counts revealing the performance of algorithm´s confusion regarding the true classifications. The observation of all the four performance measures lead to infer that the Decision Tree outperforms than other classification methods.