Fig. 1
From: Assessment of Alzheimer-related pathologies of dementia using machine learning feature selection
Methodology for classification of dementia. The methodology for the classification of dementia followed three stages: design, implementation and evaluation. First, we pre-processed and assessed feature-feature correlation after acquiring access to neuropathology and clinical data from CFAS. We then applied feature ranking methods to rank and filter all neuropathology features. Next, classifiers benchmarked with different subsets of features were selected according to their rankings. Finally, we compared cases that were consistently misclassified and evaluated brain attributes associated with these cases in order to improve machine learning