Skip to main content
. 2018 Feb 19;12:38. doi: 10.3389/fnins.2018.00038

Figure 2.

Figure 2

Classification results of three classification tasks. (A,B,C) Represents the research process of the hyper parameter k for AAL classification, Brainnetome classification and spDCM classification, respectively. For AAL classification and Brainnetome classification, the step of search is 10 while the step of search is 3 for spDCM classification considering the number of features and the speed of computation. (D) Represents ROC curve of the linear SVM classifiers in the three classification tasks. The gray dashed represents the chance level of a random classifier. (E) Represents the accuracy, recall, specificity, and f1 of linear SVM classifier in the three classification tasks. SVM: support vector machine; ROC: receiver operating characteristic.