Table 6.

Overall classification performance (accuracy, in %) for Support Vector Machine (SVM) algorithms, with 600 ms−10% overlap windows.

Validation	W	UA	T	A	L		W+A	W+L	W+T	W+UA	UA+A	UA+L	UA+T
Stage 1	94.5	$96.4$	93.5	$92.4$	93.5		$96.4$	96.5	97.0	96.8	96.8	97.4	$97.6$
Stage 2	89.3	91.0	86.8	$83.2$	87.5		92.0	92.0	92.6	92.2	92.2	$93.0$	$93.0$

Data from individual or pairs of IMUs were used as input. Stage 1= 5-fold Cross-Validation—Training dataset; Stage 2 = “leave one subject out” (LOSO) on workout dataset. Green bold numbers = best performance; red bold numbers = worst performance. W, Wrist; UA, Upper Arm; T, Thigh; A, Ankle; L, Lumbar Segment.