[Retracted] Improving Recognition of Overlapping Activities with Less Interclass Variations in Smart Homes through Clustering-Based Classification

Table 5

Confusion matrix on the Aruba dataset without activities balancing using threefold in combination of fuzzy C-means and ANN.

Acts	Slp	Tlt	MP	Rlx	HK	Eat	WD	LH	EH	WK	Res
Slp	97.9			2.1
Tlt	2.7	97.2
MP			95.2				3.1		1.7
Rlx	1.8			97.6					0.6
HK				5.1	80.0	9.7			5.2
Eat			5.8	2.0		92.2
WD			43.9	7.1			50.0
LH								55.0	45.0
EH								9.2	90.8
WK				8.4						91.6
Res						1.0				18.0	81.0

The columns represent the predicted activities, while the rows represent the actual activities. The performance of overlapping activities is highlighted in bold. Key. acts: activities, tlt: bed to toilet, eat: eating, EH: enter home, HK: housekeeping, LH: leave home, MP: meal preparation, Rlx: relax, Res: resperate, Slp: sleeping, WD: wash dishes, and WK: work.