Figure 2From: A new method for identifying bivariate differential expression in high dimensional microarray data using quadratic discriminant analysisDegradation of the performance for different classifiers. Performance of SVM with polynomial kernel (brown curve), Random Forests (green curve), QDA (black curve) and Adaboost (gray curve), estimated by 10-fold cross validation (cv), as an increasingly number of noisy features is added to the patterns defined in scenarios 1, 2, 3 and 4 (left to right and top to bottom).Back to article page