From: Feature selection for elderly faller classification based on wearable sensors
Method | Feature-Selection Subset Output | Subset # |
---|---|---|
Relief-F | Insoles: Impulse I3, I6, and I7 Head: Maximum, mean, and standard deviation posterior acceleration Maximum, mean, and standard deviation anterior acceleration Mean superior acceleration | 1 |
Relief-F | Pelvis: AP ratio of even to odd harmonics Maximum, mean, and standard deviation left acceleration Left Shank: ML Lyapunov exponent | 2 |
Relief-F | Head: Vertical ratio of even to odd harmonics Mean and standard deviation posterior acceleration Pelvis: Maximum and standard deviation left acceleration | 3 |
Relief-F | Insole: Impulse I1, I3, I4, I6, and I7 Pelvis: ML FFT first quartile AP Lyapunov exponent Maximum, mean, and standard deviation left acceleration | 4 |
Relief-F | Head: ML and vertical FFT first quartile Vertical ratio of even to odd harmonics ML Lyapunov exponent Maximum, mean, and standard deviation right acceleration Maximum, mean, and standard deviation posterior acceleration Maximum, mean, and standard deviation anterior acceleration Maximum and mean superior acceleration | 5 |
Relief-F | Pelvis: ML FFT first quartile AP ratio of even to odd harmonics AP, ML, and vertical Lyapunov exponent Maximum, mean, and standard deviation left acceleration Maximum and standard deviation inferior acceleration | 6 |
Relief-F | Insole: Impulse I3, I6, and I7 Head: Maximum, mean, and standard deviation posterior acceleration Pelvis: ML FFT first quartile AP Lyapunov exponent Maximum and mean left acceleration | 7 |
Relief-F | Pelvis: ML Lyapunov exponent Maximum, mean, and standard deviation left acceleration Left Shank: ML Lyapunov exponent Maximum and standard deviation left acceleration Maximum and standard deviation superior acceleration Right Shank: AP and vertical ratio of even to odd harmonics AP, ML, and vertical Lyapunov exponent Mean anterior acceleration | 8 |
CFS/FCBF | Pelvis: Standard deviation left acceleration | 9 |