[Retracted] Effect Measurement for Human Motion Rehabilitation Training Using Improved Deep Reinforcement Learning and IoT Networks

Table 1

Comparison of the key frame extraction error rate (%).

Number of iterations	Proposed algorithm	Literature [3] algorithm	Literature [4] algorithm	Literature [5] algorithm	Literature [6] algorithm	Literature [7] algorithm
1	1.25	5.66	10.24	17.62	20.36	23.64
2	1.66	5.89	15.61	8.61	24.55	18.99
3	1.36	7.91	12.83	15.14	15.63	18.47
4	1.54	8.62	15.97	16.55	17.18	15.99
Average value	1.45	7.02	13.66	14.48	19.43	19.27