From: Immediate word recall in cognitive assessment can predict dementia using machine learning techniques
Cross-validation accuracy (%) | Accuracy (on data split) (%) | Precision | f1-score | Sensitivity | Specificity | ||
---|---|---|---|---|---|---|---|
Dementia | Normal | ||||||
A. Scenario 2A: on the sample person’s responses | |||||||
KNN | 92.00 | 93.00 | 0.74 | 0.94 | 0.92 | 0.37 | 0.99 |
Decision tree | 90.00 | 92.00 | 0.53 | 0.94 | 0.91 | 0.42 | 0.96 |
Random forest | 92.00 | 93.00 | 0.62 | 0.94 | 0.92 | 0.42 | 0.98 |
Artificial neural networks | 92.00 | 93.00 | 0.65 | 0.95 | 0.92 | 0.44 | 0.98 |
B. Scenario 2B: on the proxy’s responses | |||||||
KNN | 71.00 | 69.00 | 0.50 | 0.70 | 0.58 | 0.02 | 0.99 |
Decision tree | 72.00 | 69.00 | 0.50 | 0.71 | 0.61 | 0.08 | 0.96 |
Random forest | 72.00 | 69.00 | 0.50 | 0.71 | 0.62 | 0.10 | 0.96 |
Artificial neural networks | 70.00 | 70.00 | 0.75 | 0.70 | 0.60 | 0.05 | 0.99 |
C. Scenario 2C: on the sample person’s and proxy’s responses | |||||||
KNN | 77.00 | 78.00 | 0.96 | 0.70 | 0.77 | 0.60 | 0.97 |
Decision tree | 77.00 | 76.00 | 0.95 | 0.68 | 0.75 | 0.56 | 0.97 |
Random forest | 77.00 | 78.00 | 0.96 | 0.69 | 0.77 | 0.59 | 0.97 |
Artificial neural networks | 76.00 | 78.00 | 0.96 | 0.69 | 0.77 | 0.59 | 0.97 |