Fig. 14From: A graphical heuristic for reduction and partitioning of large datasets for scalable supervised trainingDifference in prediction accuracies from LIBSVM for the approachesBack to article page