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

Table 8

Performance of ML (accuracy) in the inner and outer loops for each of the 10 rounds (averaged across 100 iterations). Results are reported for CDR = 1 versus CDR = 0, CDR = 0.5 versus CDR = 0, and CDR = 1 versus CDR = 0.5 using a linear kernel and the feature reduction guided by the neuropsychologists.

Round	CDR = 1 versus CDR = 0		CDR = 0.5 versus CDR = 0		CDR = 1 versus CDR = 0.5
	Inner loop accuracy	Outer loop accuracy	Inner loop accuracy	Outer loop accuracy	Inner loop accuracy	Outer loop accuracy
1
2
3
4
5
6
7
8
9
10
Total

averaged across 100 iterations; averaged across 100 iterations and 10 rounds.