Table 4.

Performance metrics of convolutional long short-term memory for chewing detection with multiple combinations of sensor data. Precision, recall, and F1-score are calculated for the eating class.

Sensor combination	Precision	Recall	F1-score
Temple	0.83	0.84	0.83
Cheek	0.92	0.88	0.90
Temple+cheek	0.92 a	0.89	0.91
Left temple+left cheek	0.9	0.86	0.88
Right temple+right cheek	0.89	0.89	0.89

^aThe selected combination shows the best results based on the F₁-score.