Click on “Download PDF” for the PDF version or on the title for the HTML version.


If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Comparing Adaptive Neuro-Fuzzy Inference System (ANFIS) to Partial Least-Squares (PLS) method for Simultaneous Prediction of Multiple Soil Properties

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  Paper number  033144,  2003 ASAE Annual Meeting . (doi: 10.13031/2013.15017) @2003
Authors:   KyeongHwan Lee, Naiqian Zhang, Sanjoy Das
Keywords:   Soil water content, Soil salinity, Adaptive neuro-fuzzy inference system (ANFIS), Partial least squares (PLS), Principal component analysis (PCA), Presision agriculture

Frequency-response data of soil at different water content and salinity levels were measured using the four-electrode Wenner array method in the frequency range of 1 Hz to 1 MHz. Partial least-squares (PLS) and adaptive neuro-fuzzy inference system (ANFIS) models were developed to simultaneously predict water content and salinity. R-square values of the PLS model for predicting water content and salinity were 0.8228 and 0.6354, respectively. The best prediction ANFIS model was found when three principal components (PCs) were used as the inputs of the model. R-square values obtained using this model reached 0.9756 and 0.9413 for water content and salinity, respectively. This study showed that both models have the potential to predict multiple soil properties simultaneously, and that the ANFIS model has a better predicting ability compared to the PLS model.

(Download PDF)    (Export to EndNotes)