Estimating topsoil texture fractions by digital soil mapping - a response to the long outdated soil map in the Philippines

Table 3 Cross validation of models by MLR applying different predictors selection methods

Texture fraction	Selection method	Predictors	MAE	RMSE	R²	Residuals Normality statistics
Texture fraction	Selection method	Predictors	MAE	RMSE	R²	Skewness	Kurtosis	^aLilliefors
Clay
	Forward	4	0.324	0.409	0.2871	−0.252	3.201	0.9187
	Backward	7	0.319	0.406	0.2974	−0.223	3.276	0.8696
	Stepwise	4	0.324	0.409	0.2871	−0.252	3.201	0.9187
Sand
	Forward	4	0.294	0.365	0.1984	0.080	2.946	0.7153
	Backward	6	0.293	0.358	0.2298	0.111	2.865	0.3559
	Stepwise	4	0.294	0.365	0.1984	0.080	2.946	0.7153
Silt
	Forward	5	0.438	0.566	0.2343	−0.670	3.616	0.0833
	Backward	5	0.435	0.565	0.2363	−0.563	3.496	0.0618
	Stepwise	4	0.443	0.569	0.2246	−0.513	3.372	0.2074

ISSN: 2468-2039