Table 2.

The comparative performance metrics for the four variants of the KIDS framework across two problem domains.

Performance metric	KIDS (I)	KIDS (II)	KIDS (III)	KIDS (IV)
Problem domain 1: sleep posture change detection
F${}_1$-score	0.91	0.90	0.97	0.99
Se	0.93	0.91	0.95	1.00
PPV	0.89	0.89	0.98	0.98
Problem domain 2: temporal segmentation of postural inactivity
R	0.63	0.78	0.81	0.96

The metrics were averaged over five participants (P1–P5). The variants encompassed: KIDS with UMAP sans the enhancement step (I), KIDS with UMAP coupled with the enhancement step (II), KIDS with ADR excluding the enhancement step (III), and KIDS amalgamating ADR with the enhancement step (IV). The metrics unveiled insights into the efficacy of the variants in addressing sleep posture change detection (Problem Domain 1) and the temporal segmentation of postural inactivity (Problem Domain 2). Notably, the metrics demonstrate the superiority of the simultaneous use of ADR and the enhancement step, as evidenced by the bolded scores indicating the highest performance across each metric.