Fig. 2
From: Sparse feature selection for classification and prediction of metastasis in endometrial cancer
Values of the discriminant function on the training cohort of 86 tumors. Negative values of the discriminant correspond to labelling the tumor as node-negative, while positive values of the discriminant correspond to labelling the tumor as node-positive. The 43 node-negative tumors are on the left side of the plot, and the 43 node-positive tumors are on the right side of the plot. It can be seen that the discriminant values of all node-negative tumors are negative, and that the discriminant values of all node-positive tumors are positive. Thus the classifier achieves 100% accuracy on the training cohort