EURASIP Journal on Wireless Communications and Networking

Table 5 The prediction performance of nine different hybrid models

From: Using improved support vector regression to predict the transmitted energy consumption data by distributed wireless sensor network

Models	MAE	MSE	RMSE	R²(%)
RF + Lars	0.100304	0.013272	0.115202	80.791
RF + stepwise	0.103614	0.013119	0.114538	81.012
RF + Boruta	0.089666	0.012335	0.111064	82.146
GBR + Lars	0.099679	0.012001	0.109548	82.63
GBR + Boruta	0.095545	0.010855	0.104188	84.288
GBR + stepwise	0.089575	0.010729	0.103579	84.472
SVR + stepwise	0.079161	0.009541	0.097676	86.191
SVR + Lars	0.077012	0.009265	0.096255	86.590
SVR + Boruta	0.074670	0.006505	0.080651	90.585

Back to article page