A comparison of support vector machines, artificial neural network and classification tree for identifying soil texture classes in southwest China

Abstract The variability of soil properties plays a critical role in soil and water conversation engineering. In this study, different machine learning techniques were applied to identify the soil texture classes based on a set of terrain parameters in a small mountainous watershed located in the core areas of Three Gorges of Yangtze River, southwest China. For this, the support vector machines (SVMs) with polynomial and Gaussian radius basis functions, artificial neural network, and classification tree methods were compared. The most commonly used performance measures including overall accuracy, kappa index, receiver operating characteristics (ROC), and area under the ROC curve (AUC) were employed to evaluate the accuracy of the models for classification. The observed results showed a better performance under SVMs than under artificial neural network and classification tree algorithms. Moreover, SVM with polynomial function (SVM-poly) worked slightly better than SVM with Gaussian radius basis function. The overall accuracy, kappa statistic, and AUC of SVM-poly were 0.943, 0.79, and 0.944, respectively. Meanwhile, the classification accuracy was 0.794 for clay, 0.992 for loam, and 0.661 for sand under SVM-poly. Elevation, terrain classification index for lowlands, and flow path length were the most important terrain indicators affecting the variation in the soil texture class in the study area. These results showed that the support vector machines are feasible and reliable in the identification of soil texture classes.

[1]  Timothy A. Quine,et al.  THE RELATIVE CONTRIBUTION OF SOIL TILLAGE AND OVERLAND FLOW EROSION TO SOIL REDISTRIBUTION ON AGRICULTURAL LAND , 1996 .

[2]  D. K. Cassel,et al.  Soil-Landscape Relationships in a Closed Drainage System 1 , 1974 .

[3]  Shamsollah Ayoubi,et al.  Prediction of Soil Enzymes Activity by Digital Terrain Analysis: Comparing Artificial Neural Network and Multiple Linear Regression Models , 2012 .

[4]  Anne Gobin,et al.  Soil-landscape modelling to quantify spatial variability of soil texture. Modelling of transport pro , 1999 .

[5]  M. Kovacevic,et al.  Soil type classification and estimation of soil properties using support vector machines , 2010 .

[6]  J. C. Obi,et al.  Development of pedotransfer functions for coastal plain soils using terrain attributes , 2014 .

[7]  Thomas C. Edwards,et al.  Machine learning for predicting soil classes in three semi-arid landscapes , 2015 .

[8]  Gary A. Peterson,et al.  Soil Attribute Prediction Using Terrain Analysis , 1993 .

[9]  Wei Wu,et al.  Effects of land use and topographical attributes on soil properties in an agricultural landscape , 2011 .

[10]  Gerard B. M. Heuvelink,et al.  Sampling for validation of digital soil maps , 2011 .

[11]  Wei Wu,et al.  Assessing effects of digital elevation model resolutions on soil-landscape correlations in a hilly area , 2008 .

[12]  Z. H. Zhang,et al.  Effect of terrace forms on water and tillage erosion on a hilly landscape in the Yangtze River Basin, China , 2014 .

[13]  Michael Bock,et al.  System for Automated Geoscientific Analyses (SAGA) v. 2.1.4 , 2015 .

[14]  B. Minasny,et al.  Comparing data mining classifiers to predict spatial distribution of USDA-family soil groups in Baneh region, Iran , 2015 .

[15]  Wei-Yin Loh,et al.  Classification and regression trees , 2011, WIREs Data Mining Knowl. Discov..

[16]  John H. Grove,et al.  Soil-landscape modeling across a physiographic region : Topographic patterns and model transportability , 2006 .

[17]  Paul E. Gessler,et al.  Modeling Soil–Landscape and Ecosystem Properties Using Terrain Attributes , 2000 .

[18]  Jin Zhang,et al.  An overview and comparison of machine-learning techniques for classification purposes in digital soil mapping , 2016 .

[19]  Fan-Rui Meng,et al.  Predict soil texture distributions using an artificial neural network model , 2009 .

[20]  J. E. Foss,et al.  Testing the Accuracy of Field Textures 1 , 1975 .

[21]  J. Ross Quinlan,et al.  Induction of Decision Trees , 1986, Machine Learning.

[22]  Stephen Fityus,et al.  The development of a residual soil profile from a mudstone in a temperate climate , 2004 .

[23]  I. Moore,et al.  Digital terrain modelling: A review of hydrological, geomorphological, and biological applications , 1991 .

[24]  R. Kerry,et al.  Digital mapping of soil organic carbon at multiple depths using different data mining techniques in Baneh region, Iran , 2016 .

[25]  Qing-Kai Sheng,et al.  Prediction of soil organic matter using artificial neural network and topographic indicators in hilly areas , 2013, Nutrient Cycling in Agroecosystems.

[26]  Yakov A. Pachepsky,et al.  Hydropedology and pedotransfer functions , 2006 .

[27]  Ahmad Jalalian,et al.  Relationships between soil depth and terrain attributes in a semi arid hilly region in western Iran , 2013, Journal of Mountain Science.

[28]  S. Thien.  A flow diagram for teaching texture‐by‐feel analysis , 1979 .

[29]  Murray K. Clayton,et al.  Potential terrain controls on soil color, texture contrast and grain-size deposition for the original catena landscape in Uganda , 2004 .

[30]  Kilian Stoffel,et al.  Theoretical Comparison between the Gini Index and Information Gain Criteria , 2004, Annals of Mathematics and Artificial Intelligence.

[31]  B. Ehret.  Pattern recognition of geophysical data , 2010 .

[32]  B. Huwe,et al.  Uncertainty in the spatial prediction of soil texture: Comparison of regression tree and Random Forest models , 2012 .

[33]  Donald W. Marquaridt.  Generalized Inverses, Ridge Regression, Biased Linear Estimation, and Nonlinear Estimation , 1970 .

[34]  H. Jenny,et al.  Factors of Soil Formation , 1941 .

[35]  Wolfgang Wilcke,et al.  Soils Along the Altitudinal Transect and in Catchments , 2008 .

[36]  Annette Freibauer,et al.  Field-based soil-texture estimates could replace laboratory analysis , 2016 .

[37]  Igor V. Florinsky,et al.  Prediction of soil properties by digital terrain modelling , 2002, Environ. Model. Softw..

[38]  Alfredo Huete,et al.  A comparison of soil scientist estimations and laboratory determinations of some Arizona soil properties , 1986 .

[39]  Tereza Zádorová,et al.  Colluvial soils as a soil organic carbon pool in different soil regions , 2015 .

[40]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[41]  Christopher J. C. Burges,et al.  A Tutorial on Support Vector Machines for Pattern Recognition , 1998, Data Mining and Knowledge Discovery.