w*n
1 楼
From: International Journal of Hydrology Science and Technology
Title: Comparison between support vector machine and nonlinear regression
for predicting saturated hydraulic conductivity
Required by Mar 19, 2018.
Abstract
Saturated hydraulic conductivity (Ks) is playing an important role in
irrigation, and drainage. The aim of the study was to validate Pedotransfer
Functions (PTFs) using non-linear regression (NLR) and support vector
machine (SVM) for estimation Ks. Moreover, selecting the best predictor
variables used for determination of PTFs. Six classes of PTFs were proposed
based on soil physical properties using both NLR and SVM were: Ks-1 (Clay +
Silt+ Sand), Ks-2 (Clay +Silt +Sand+ Bulk density), Ks-3 (Clay+ silt+ sand+
Organic matter), Ks-4 (Clay +Silt +Sand+ Bulk density + Organic matter), Ks-
5 (Clay+ Bulk density + Organic matter) and Ks-6 (Clay+ Sand+ Bulk density +
Organic matter), respectively. Ks was measured by a constant head method
and predicted by proposed PTFs based on NRL and SVM. Generally, RMSE for the
SVM was less than RMSE for NLR for predicting Ks for all PTFs classes. The
best proposed class developed by NLR was Ks-2, RMSE= 2.72×10-6 m/s. While
the best proposed class developed by SVM was Ks-3, RMSE= 8.86×10-7m/s. SVM
is greater than NLR in the performance for predicting Ks using PTFs. SVM
modeling approach can be used for estimations Ks with the accuracy
comparable to one of the physically based models.
Keywords: Soil hydraulic properties; nonlinear regression model; Support
vector machine; Pedotransfer functions (PTFs); Mathematical models.
Introduction
PM Me if you are interested.
Thanks.
Title: Comparison between support vector machine and nonlinear regression
for predicting saturated hydraulic conductivity
Required by Mar 19, 2018.
Abstract
Saturated hydraulic conductivity (Ks) is playing an important role in
irrigation, and drainage. The aim of the study was to validate Pedotransfer
Functions (PTFs) using non-linear regression (NLR) and support vector
machine (SVM) for estimation Ks. Moreover, selecting the best predictor
variables used for determination of PTFs. Six classes of PTFs were proposed
based on soil physical properties using both NLR and SVM were: Ks-1 (Clay +
Silt+ Sand), Ks-2 (Clay +Silt +Sand+ Bulk density), Ks-3 (Clay+ silt+ sand+
Organic matter), Ks-4 (Clay +Silt +Sand+ Bulk density + Organic matter), Ks-
5 (Clay+ Bulk density + Organic matter) and Ks-6 (Clay+ Sand+ Bulk density +
Organic matter), respectively. Ks was measured by a constant head method
and predicted by proposed PTFs based on NRL and SVM. Generally, RMSE for the
SVM was less than RMSE for NLR for predicting Ks for all PTFs classes. The
best proposed class developed by NLR was Ks-2, RMSE= 2.72×10-6 m/s. While
the best proposed class developed by SVM was Ks-3, RMSE= 8.86×10-7m/s. SVM
is greater than NLR in the performance for predicting Ks using PTFs. SVM
modeling approach can be used for estimations Ks with the accuracy
comparable to one of the physically based models.
Keywords: Soil hydraulic properties; nonlinear regression model; Support
vector machine; Pedotransfer functions (PTFs); Mathematical models.
Introduction
PM Me if you are interested.
Thanks.