Table 7.

Accuracy (ACC), precision (PPV, Positive Predictive Value) and sensitivity (TPR, True Positive Rate) for each functional fitness movement, related to the cubic kernel SVM algorithm with a 600 ms length windows and 10% overlap.

	W			UA			T			A			L
	ACC	PPV	TPR	ACC	PPV	TPR	ACC	PPV	TPR	ACC	PPV	TPR	ACC	PPV	TPR
C&J	$96.4$	85.7	$75.9$	96.2	$86.7$	72.9	93.0	65.1	62.9	$90.9$	$53.6$	$55.7$	93.2	67.6	58.5
AS	$96.8$	81.3	$48.1$	97.3	$78.5$	63.7	97.5	81.7	65.9	97.6	$82.6$	65.2	$97.8$	82.1	$71.1$
BJ	98.6	83.9	76.8	$99.1$	$90.9$	$83.7$	96.9	61.3	45.3	$95.4$	$36.5$	$28.1$	98.1	87.9	57.1
BP	97.0	$88.3$	$58.2$	$97.9$	83.6	$81.5$	97.5	85.3	69.8	97.1	79.5	69.2	$97.0$	$78.9$	68.0
TRANS	89.7	90.2	$96.8$	$91.5$	$92.7$	96.3	88.6	90.6	94.7	$85.3$	$89.0$	$91.8$	89.0	90.5	95.5

Data from individual IMUs were used as input. Classification performance is reported as %. For sake of clarity, only results from the most stringent validation (Stage 2 = “leave one subject out” on test dataset) are reported. Green bold numbers = best performance; red bold numbers = worst performance. W, Wrist; UA, Upper Arm; T, Thigh; A, Ankle; L, Lumbar Segment.