Table 4 Summary of the balanced accuracies of the machine learning classifiers
Train-Test Split | PRO | Group | Mean Accuracy | Highest Accuracy | ||
|---|---|---|---|---|---|---|
PF | MF | PF | MF | |||
Intersubject CV method | Macros | All | 50.69 ± 2% | 49.70 ± 1% | 52.98% | 51.69% |
Walking Volume | 47.35 ± 1% | 49.56 ± 0% | 48.85% | 49.95% | ||
Non-walking Volume | 49.12 ± 3% | 49.62 ± 1% | 50.71% | 51.39% | ||
Pattern | 47.91 ± 1% | 50.64 ± 1% | 49.73% | 51.72% | ||
Vector Magnitude | 52.73 ± 1% | 49.99 ± 1% | 54.47% | 50.60% | ||
Variability | 49.29 ± 1% | 51.10 ± 1% | 50.02% | 51.86% | ||
Micros | All | 46.62 ± 3% | 48.36 ± 1% | 50.09% | 50.42% | |
Pace | 43.62 ± 2% | 49.11 ± 2% | 45.53% | 51.39% | ||
Variability | 47.40 ± 2% | 47.37 ± 2% | 49.85% | 49.02% | ||
Rhythm | 47.58 ± 2% | 48.50 ± 2% | 49.83% | 50.86% | ||
Asymmetry | 48.85 ± 1% | 49.82 ± 1% | 49.84% | 51.81% | ||
Postural Control | 48.09 ± 2% | 50.55 ± 1% | 50.72% | 51.89% | ||
Intrasubject CV method | Macros | All | 53.44 ± 1% | 53.21 ± 1% | 54.54% | 54.53% |
Micros | All | 56.13 ± 0% | 54.70 ± 1% | 56.90% | 55.74% | |