a nonparametric model by using k-nearest neighbor technique for predicting soil saturated hydraulic conductivity

abstract saturated hydraulic conductivity (ks) is needed for many studies related to water and solute transport, but often cannot be measured because of practical and/or cost-related reasons. nonparametric approaches are being used in various fields to estimate continuous variables. one type of the nonparametric lazy learning algorithms, a k-nearest neighbor (k-nn) algorithm, was introduced and tested to estimate saturated hydraulic conductivity (ks) from other soil properties including soil textural fractions, ec, ph, sp, oc, tnv, ρs and ρb. a number of 10 nearest neighbors, based on cross validation technique were selected to perform saturated hydraulic conductivity prediction from 151 soil sample attributes. the nonparametric k-nn technique performed mostly equally well, in terms of pearson correlation coefficient (r=0.801), modeling efficiency (ef=0.65), root-mean-squared errors (rmse=71.15) maximum error (me=120.47), coefficient of determination (cd=1.32) and coefficient of residual mass (crm=-0.046) statistics. it can be concluded that the k-nn technique is a competitive alternative to other techniques such as pedotransfer functions (ptfs) to estimate saturated hydraulic conductivity.