Optimizing Neuropsychological Assessments for Cognitive, Behavioral, and Functional Impairment Classification: A Machine Learning Study

Figure 3

Highest accuracy of classification as a function of the configuration (combination of input features) in the inner loop of the nested CV. Performances are reported for CDR = 1 versus CDR = 0 (blue), CDR = 0.5 versus CDR = 0 (red), and CDR = 1 versus CDR = 0.5 (green), using the features chosen by the neuropsychologists. Results were obtained as average across all rounds (10) and iterations (100).