Comparing Neural Networks, Linear and Nonlinear Regression Techniques to Model Penetration Resistance

Hossein BAYAT; Mohammad Reza NEYSHABOURI; Mohammad Ali HAJABBASI

Makale detayları

Comparing Neural Networks, Linear and Nonlinear Regression Techniques to Model Penetration Resistance

Yazarlar:

Hossein BAYAT | Mohammad Reza NEYSHABOURI | Mohammad Ali HAJABBASI

Yayıncı:

TUBITAK

Dergi:

Turkish Journal of Agriculture and Forestry 2008 Vol.32 Issue 5, pp.425-433

Belge Tipi:

Makale

Yayın Tarihi:

4/2008

Belge Dili:

Türkçe

Document ID (Org No.):

2000808595

Sıklık:

İki Aylık

eISSN:

1303-6173

pISSN:

1300-011X

Anahtar kelimeler:

Özet:

 

Penetration resistance (PR) is an important property of soils, and can be expressed as cone index (CI). Because of high variability, there are no accurate and representative PR data in most cases. Variable PR is considerably affected by gravimetric soil water content (GWC) and bulk density (BD). In this study, artificial neural networks (ANNs) were used to simulate relationship between BD, GWC, and CI. A data set of 381 samples was collected from 2 study sites, Hamadan and Maragheh. Pedotransfer functions (PTFs) were developed using ANNs and linear and nonlinear regression models to predict CI for the combined data set and each data set separately. For the combined and Hamadan data sets, ANNs produced a greater correlation coefficient (R = 0.85) and lower root mean square error (RMSE) compared with the linear regression model (R = 0.70). For the Maragheh data set, however, the regression model yielded better results. Introducing TP and relative saturation (Qv/TP) into the models improved the prediction of CI. The results further showed that ANN models performed better than nonlinear regression models. Therefore, ANNs were recognized as powerful tools to predict CI by BD, GWC, TP, and Qv/TP as the independent variables under the very diverse conditions of the soils and treatments employed.

 

Penetration resistance (PR) is an important property of soils, and can be expressed as cone index (CI). Because of high variability, there are no accurate and representative PR data in most cases. Variable PR is considerably affected by gravimetric soil water content (GWC) and bulk density (BD). In this study, artificial neural networks (ANNs) were used to simulate relationship between BD, GWC, and CI. A data set of 381 samples was collected from 2 study sites, Hamadan and Maragheh. Pedotransfer functions (PTFs) were developed using ANNs and linear and nonlinear regression models to predict CI for the combined data set and each data set separately. For the combined and Hamadan data sets, ANNs produced a greater correlation coefficient (R = 0.85) and lower root mean square error (RMSE) compared with the linear regression model (R = 0.70). For the Maragheh data set, however, the regression model yielded better results. Introducing TP and relative saturation (Qv/TP) into the models improved the prediction of CI. The results further showed that ANN models performed better than nonlinear regression models. Therefore, ANNs were recognized as powerful tools to predict CI by BD, GWC, TP, and Qv/TP as the independent variables under the very diverse conditions of the soils and treatments employed.

Makale detayları

Comparing Neural Networks, Linear and Nonlinear Regression Techniques to Model Penetration Resistance

Yazarlar:

Hossein BAYAT | Mohammad Reza NEYSHABOURI | Mohammad Ali HAJABBASI

Yayıncı:

TUBITAK

Dergi:

Turkish Journal of Agriculture and Forestry 2008 Vol.32 Issue 5, pp.425-433

Belge Tipi:

Makale

Yayın Tarihi:

4/2008

Belge Dili:

Türkçe

Document ID (Org No.):

2000808595

Sıklık:

İki Aylık

eISSN:

1303-6173

pISSN:

1300-011X

Artificial neural networks

bulk density

cone index

regression models

water content

Sayı

Makale detayları

Comparing Neural Networks, Linear and Nonlinear Regression Techniques to Model Penetration Resistance

Yazarlar: Hossein BAYAT | Mohammad Reza NEYSHABOURI | Mohammad Ali HAJABBASI

Yayıncı: TUBITAK

Dergi: Turkish Journal of Agriculture and Forestry 2008 Vol.32 Issue 5, pp.425-433

Belge Tipi: Makale

Yayın Tarihi: 4/2008

Belge Dili: Türkçe

Document ID (Org No.): 2000808595

Sıklık: İki Aylık

eISSN: 1303-6173

pISSN: 1300-011X

Artificial neural networks

bulk density

cone index

regression models

water content

Sayı

Yazarlar:

Hossein BAYAT | Mohammad Reza NEYSHABOURI | Mohammad Ali HAJABBASI

Yayıncı:

TUBITAK

Dergi:

Turkish Journal of Agriculture and Forestry 2008 Vol.32 Issue 5, pp.425-433

Belge Tipi:

Makale

Yayın Tarihi:

4/2008

Belge Dili:

Türkçe

Document ID (Org No.):

2000808595

Sıklık:

İki Aylık

eISSN:

1303-6173

pISSN:

1300-011X