Table 6.
Classification results for binary classification models G0. versus {G1+G2+}.
| Dataset and classifier | AUCa | Accuracy | Recall | Precision | F1 | ||
| Wearable |
|
|
|
|
|
||
|
|
LRb | 0.680 (+/-0.323) | 0.622 (+/-0.215) | 0.604 (+/-0.222) | 0.553 (+/-0.289) | 0.552 (+/-0.250) | |
|
|
DTc | 0.725 (+/-0.183) | 0.757 (+/-0.139) | 0.725 (+/-0.183) | 0.676 (+/-0.278) | 0.670 (+/-0.230) | |
|
|
RFd | 0.757 (+/-0.250) | 0.784 (+/-0.236) | 0.757 (+/-0.250) | 0.729 (+/-0.325) | 0.720 (+/-0.294) | |
| RAI-HCe |
|
|
|
|
|
|
|
|
|
LR | 0.840 (+/-0.236) | 0.730 (+/-0.213) | 0.689 (+/-0.249) | 0.644 (+/-0.304) | 0.648 (+/-0.276) | |
|
|
DT | 0.856 (+/-0.153) | 0.865 (+/-0.132) | 0.856 (+/-0.153) | 0.868 (+/-0.181) | 0.836 (+/-0.179) | |
|
|
RF | 0.858 (+/-0.160) | 0.865 (+/-0.145) | 0.858 (+/-0.160) | 0.870 (+/-0.186) | 0.836 (+/-0.189) | |
| Wearable + RAI-HC |
|
|
|
|
|
||
|
|
LR | 0.743 (+/-0.251) | 0.784 (+/-0.224) | 0.766 (+/-0.232) | 0.778 (+/-0.255) | 0.755 (+/-0.246) | |
|
|
DT | 0.842 (+/-0.229) | 0.865 (+/-0.192) | 0.851 (+/-0.213) | 0.886 (+/-0.202) | 0.849 (+/-0.214) | |
|
|
RF | 0.865 (+/-0.125) | 0.865 (+/-0.132) | 0.865 (+/-0.125) | 0.908 (+/-0.094) | 0.853 (+/-0.139) | |
aAUC: area under the receiver operating characteristic curve.
bLR: logistic regression.
cDT: decision tree.
dRF: random forest.
eRAI-HC: Resident Assessment Instrument-Home Care.