Table 2 Accuracy of the FAS and FMA estimation algorithms investigated in the study.

From: Enabling precision rehabilitation interventions using wearable sensors and machine learning to track motor recovery

 

RMSE

r2

FAS (0–5 points)

Random forest

0.38

0.79

FMA (0–66 points)

Linear regression (\(\widehat {\mathrm{{FAS}}}\))

7.79

0.47

Random forest

5.05

0.77

Balanced random forest

4.17

0.84

Proposed technique

3.99

0.86

  1. Root-mean-square error (RMSE) and coefficient of determination (r2) values are shown for the FAS score estimates derived using a random forest-based algorithm as well as for the four methods implemented in the study to estimate FMA scores.
Back to article page