Wearable Sensor-Based Human Activity Recognition Using Hybrid Deep Learning Techniques

Table 11

Average accuracy of different activities with five deep learning models.

ID	CNN (%)	LSTM (%)	CNN-BLSTM (%)	CNN-GRU (%)	CNN-LSTM (%)
A1	97.50	97.70	97.41	99.75	99.03
A2	97.25	97.10	95.65	98.99	97.78
A3	95.60	97.15	100	96.57	98.86
A4	91.26	90.26	91.96	81.99	90.76
A5	90.80	90.80	84.74	92.48	93.09
A6	99.67	98.58	100	99.78	99.78
A7	76.47	64.86	44.44	77.78	94.44
A8	100	66.67	66.67	50.00	100
A9	63.83	69.39	62.07	48.00	76.00
A10	84.85	70.27	80.00	52.94	82.35
A11	72.50	69.33	65.00	71.79	82.05
A12	83.30	70.27	70.59	55.56	88.89