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

Table 3

Performance evaluation metrics on the Aruba dataset without activities balancing using threefold cross-validation.

Dataset	Cross-validation	Clustering method	Classification method	Precision (%)	Recall (%)	F score [0,1]	Accuracy (%)
Aruba	Threefolds	Fuzzy C-means [35]	ANN [45]	84.50	84.70	0.84	84.70
			ET-KNN [47]	80.60	80.80	0.80	80.80
			KNN [48]	79.30	78.50	0.79	78.50
			SMO [46]	75.37	74.02	0.74	74.02
		Hierarchical [42]	ANN	82.80	83.80	0.82	83.80
			ET-KNN	80.80	80.70	0.80	80.80
			KNN	78.20	78.20	0.78	78.20
			SMO	76.02	75.01	0.74	76.02
		K-mean [36]	ANN	81.30	82.80	0.82	82.80
			ET-KNN	79.50	79.80	0.79	79.80
			KNN	77.50	78.20	0.78	78.20
			SMO	74.20	75.01	0.74	75.02
		DBSCAN [37]	ANN	79.20	80.80	0.80	80.80
			ET-KNN	78.30	78.80	0.78	78.80
			KNN	76.20	75.10	0.76	76.20
			SMO	74.02	74.02	0.74	74.02

The precision, recall, and accuracy are in percentages (%), while the range of F score is between [0-1] with 1 being the highest. The highest values are in bold.