Predicting Saturated Hydraulic Conductivity by Artificial Intelligence and Regression Models

Table 1

Descriptive statistics of the datasets used for training and testing (ANN, ANFIS, and MLR).

Variable	Units	Training data (𝑁 = 130)				Testing data (𝑁 = 45)
		Max.	Min.	Mean	S.D.	Max.	Min.	Mean	S.D.
𝜌𝑏	Mg m^−3	1.77	0.96	1.45	0.14	1.74	1.20	1.45	0.13
Sand	%	70.0	4.0	27.9	13.8	52.0	4.0	20.4	11.2
Silt	%	66.0	12.0	44.6	10.6	78.0	18.0	47.5	10.8
Clay	%	62.0	9.4	28.5	11.2	62.0	10.5	32.1	12.6
𝐾𝑠	cm day^−1	207.0	6.9	27.4	29.4	60.0	6.4	18.2	10.7