In the second set of 3 analyses, analyses 6–8 of Table Table1,1, all subcortical and cortical measures were used, either by merging feature sets or through Ensemble methods. For direct comparison with Ensemble methods, we carried out analysis 6 of Table Table1,1, where all subcortical and cortical measures were merged, redundant variables were removed and feature selection and classification were carried out in a manner similar to independent measures classification. In analysis 7, we employed three different input classifiers for Ensemble classification, where input classifiers were trained on subcortical volumes, cortical volumes and remaining cortical measures respectively as shown in Table Table1.1. In analysis 8, five different input classifiers trained on independent measure sets, from analyses 1–5, are used for Ensemble classification.
Do you have any questions about this protocol?
Post your question to gather feedback from the community. We will also invite the authors of this article to respond.