Assessing Dry Weight of Hemodialysis Patients via Sparse Laplacian Regularized RVFL Neural Network with L_2,1-Norm

Table 2

Comparison on existing methods via 10-fold cross-validation.

Method		squared	RMSE	Empirical cumulative distribution plot
				Highest value	Lowest value	Median value
BCM	0.9473	0.9137	1.9694	3.2235	-6.2776	-0.9863
LR	0.9403	0.9308	1.4335	4.2524	-4.4014	0.1418
ANN (BP)	0.9398	0.9295	1.4794	7.3661	-4.7447	0.1324
MKRR	0.9399	0.9289	1.5015	4.9227	-4.2604	0.1104
MKSVR	0.9412	0.9321	1.3817	4.3962	-4.1273	0.0082
RVFL	0.9389	0.9300	1.3828	6.7004	-4.3557	0.0704
SLapRVFL (our method)	0.9632	0.9501	1.3136	3.1940	-3.5066	0.1014

The results are from previous work on MKSVR [39].