THE APPLICATION OF ARTIFICIAL INTELLIGENCE METHODS FOR EVALUATION AND MODELLING OF THE SOIL PROPERTIES

Abstract

As artificial intelligence (AI) applications see wider deployment, it becomes increasingly important to study the social and societal implications of AI adoption. AI has been in existence for over six decades and has experienced AI winters and springs. The rise of super computing power and Big Data technologies appear to have empowered AI in recent years. The new generation of AI is rapidly expanding and has again become an attractive topic for research. The paper first provides a view of the history of AI through the relevant papers published in the International Journals. Examination of the soil properties like Cation Exchange Capacity (CEC) plays an important role in the study of environmental researches. The spatial and temporal variability of this property has been caused to the development of indirect methods in estimation of the soil characteristics. This paper aims to employ different AI-based methods to estimate the cation exchange capacity. One hundred and fifty soil samples are collected from different horizons of soil profiles located in the Behbahan region, Khuzestan Province, Southwest of Iran. Finally, multiple linear regression, Neuro-Fuzzy, feedforward back-propagation network, and other methods are employed to develop a pedotransfer function (PTF) for predicting soil parameters using easily measurable characteristics of clay and organic carbon. As an interesting consequence, Neuro-Fuzzy, SVM, and some others are superior to artificial neural networks and multiple linear regression in predicting soil property.