Numerical simulation and optimization method of sports teaching and training based on embedded wireless communication network

Figure 1.

Figure 2.

Figure 3.

Influence of the main components of ASPP+LSTM_

Method	Components	PAMAP2					MHEALTH
		Fβ	MAE	Fβω$$F_\beta ^\omega$$	Sm	Em	Fβ	MAE	Fβω$$F_\beta ^\omega$$	Sm	Em
M1	CNN	0.701	0.147	0.514	0.612	0.601	0.821	0.101	0.605	0.708	0.413
M2	LSTM	0.689	0.151	0.507	0.605	0.589	0.812	0.105	0.598	0.700	0.405
M3	CNN+LSTM	0.710	0.144	0.520	0.616	0.609	0.825	0.099	0.610	0.710	0.418
M4	ASPP	0.703	0.146	0.516	0.614	0.603	0.822	0.100	0.607	0.709	0.414
M5	ASPP+LSTM	0.723	0.140	0.530	0.625	0.617	0.834	0.096	0.618	0.715	0.425

Compares our model with other deep learning mainstream methods in terms of Fβ(↓)$${F_\beta }\left( \downarrow \right)$$, MAE(↓)$$MAE\left( \downarrow \right)$$, Fβω(↑)$$F_\beta ^\omega \left( \uparrow \right)$$, and  Sm(↑)$$\;{S_m}\left( \uparrow \right)$$ on two datasets_ The best result for each column is highlighted in bold_

Method	PAMAP2				MHEALTH
	Fβ	MAE	Fβω$$F_\beta ^\omega$$	Sm	Fβ	MAE	Fβω$$F_\beta ^\omega$$	Sm
AFNet [3]	0.721	0.184	0.526	0.636	0.815	0.114	0.612	0.708
DSS [7]	0.683	0.197	0.489	0.608	0.782	0.127	0.598	0.681
HRSOD [39]	0.692	0.193	0.505	0.617	0.795	0.122	0.605	0.690
FCSOD [40]	0.701	0.189	0.513	0.625	0.804	0.119	0.610	0.700
PA-KRN [36]	0.712	0.186	0.520	0.632	0.810	0.116	0.615	0.705
TSPOANe t[14]	0.716	0.185	0.523	0.634	0.813	0.115	0.618	0.707
Our	0.740	0.178	0.540	0.649	0.822	0.110	0.625	0.715

Our model is compared with 17 state-of-the-art methods in terms of  Em(↑)$$\;{E_m}\left( \uparrow \right)$$ on 2 datasets_

Method	PAMAP2	MHEALTH	Method	PAMAP2	MHEALTH
	Em			Em
AFNet [3]	0.632	0.471	CPD [35]	0.788	0.715
DSS [7]	0.624	0.586	BASNet [21]	0.763	0.728
HRSOD [39]	0.682	0.524	GCPANet [2]	0.722	0.762
FCSOD [40]	0.642	0.623	LDF [33]	0.749	0.725
PA-KRN [36]	0.628	0.608	ITSD [46]	0.792	0.781
TSPOANet [14]	0.692	0.611	MINet [18]	0.814	0.744
BRN [31]	0.715	0.644	GateNet [45]	0.826	0.791
PiCA [13]	0.754	0.672	DUCRF [37]	0.851	0.821
PoolNet [12]	0.761	0.701	Our	0.869	0.865