Multimodal Signatures of Brain Aging: From Descriptive Analyses to Machine Learning-Based Integration.

When behavioral features were combined with LFP-derived functional connectivity measures (TotCoh across frequency bands; dataset 2), classification performance showed a numerical improvement that did not reach statistical significance, yielding an overall accuracy of 0.788 (95% CI: 0.697–0.880) using a Linear Support Vector Classifier (LinearSVC) ( Figure 8 b).

also in 111,027 other papers