Application of ANFIS-based subtractive clustering algorithm in soil Cation Exchange Capacity estimation using soil and remotely sensed data. (January 2017)

• Record Type: Journal Article
• Authors: Keshavarzi, Ali; Sarmadian, Fereydoon; Shiri, Jalal; Iqbal, Munawar; Tirado-Corbalá, Rebecca; Omran, El-Sayed Ewis
• Is Part Of: Measurement. Volume 95(2017)
• Page Start: 173
• Page End: 180
• Publication Date: 2017-01
• Subjects: ANFIS -- Multi variable linear regression -- Soil Cation Exchange Capacity -- Soil properties -- Subtractive clustering; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2016.10.010
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: The conventional procedures for soil Cation Exchange Capacity ( CEC ) measurement are time consuming. MR and ANFIS models were employed to predict the soil CEC based on clay, organic carbon ( OC ) and Band 3 inputs. The results revealed that the ANFIS model had the ability to estimate soil CEC by computing easily measurable variables. Abstract: The conventional procedures for soil Cation Exchange Capacity ( CEC ) measurement are time consuming and laborious. It is also difficult to maintain stability for long-term experiments and projects. Therefore, this study aimed at comparing Adaptive Neuro-Fuzzy Inference System (ANFIS)-based subtractive clustering algorithm with different inputs combinations as well as sequential regression models for simulation of variations in soil CEC . Results showed that the corresponding values of root mean squared error ( RMSE ) and coefficient of determination ( R 2 ) between the measured and simulated CEC using the best regression equation and ANFIS models were 2.05 and 0.733, 1.35 and 0.806, respectively. Nevertheless, sensitivity analysis was conducted to determine the most and the least influential variables affecting soil CEC . Results of the present investigation showed that the ANFIS model had the ability to estimate soil CEC by computing easily measurable variables with guarantee of authenticity, reliability and reproducibility.