sEMG-Based Neural Network Prediction Model Selection of Gesture Fatigue and Dataset Optimization

Table 1

Temperature and wildlife count in the three areas covered by the study.

sEMG		MSE	RMSE	MAE
BP-G1	Train	0.0196	0.3385	0.3127
	Test	0.0085	0.3421	0.3306
BP-G4	Train	0.0339	0.2218	0.1782
	Test	0.0243	0.1979	0.1617
BP-G7	Train	0.0395	0.2468	0.2032
	Test	0.0395	0.2327	0.1911
RNN-G1	Train	0.0214	0.1354	0.0928
	Test	0.0069	0.086	0.0579
RNN-G4	Train	0.033	0.1812	0.1487
	Test	0.0261	0.1612	0.1289
RNN-G7	Train	0.04	0.1169	0.1024
	Test	0.0376	0.1107	0.0984
LSTM-G1	Train	0.0214	0.1133	0.0768
	Test	0.0069	0.0696	0.0491
LSTM-G4	Train	0.033	0.1842	0.152
	Test	0.0261	0.166	0.1331
LSTM-G7	Train	0.04	0.0763	0.06
	Test	0.0376	0.0717	0.0569
SVR-G1	Train	0.0214	0.0576	0.0439
	Test	0.0069	0.04	0.0305
SVR-G4	Train	0.033	0.1643	0.1298
	Test	0.0261	0.1488	0.1162
SVR-G7	Train	0.04	0.0482	0.0373
	Test	0.0376	0.046	0.0361