Table 6.
Sensitivity (true-positive rate, TPR), specificity (true-negative rate, TNR), and accuracy (ACC) for the faller classification using the TUG 6, SWAP 4, and SWAP 82 tasks with different number of epochs in ANN learning (averaged over number of neurons and regularisation approach).
| TUG 6 | SWAP 4 | SWAP 82 | |||||||
|---|---|---|---|---|---|---|---|---|---|
| TPR | TNR | ACC | TPR | TNR | ACC | TPR | TNR | ACC | |
| 2 epochs | 73% | 31% | 65% | 80% | 42% | 73% | 86% | 54% | 80% |
| 4 epochs | 95% | 13% | 79% | 88% | 15% | 73% | 84% | 58% | 79% |
| 20 epochs | 95% | 29% | 82% | 84% | 21% | 72% | 85% | 77% | 84% |