Prediction of organic potato yield using tillage systems and soil properties by artificial neural network (ANN) and multiple linear regressions (MLR)

Abstract

Tillage aims to prepare the soil with the adequate treatment to create the ideal and most favorable conditions for cultivation. To evaluate the effect of tillage systems on soil environment, it is mandatory to measure the modifications in physical, chemical and biological properties. In recent decades, artificial intelligence systems were used for developing predictive models to simplify, estimate and predict many farming processes. They are also employed to optimize performance and control risks. These systems have become true virtual helpers, and more so when integrated with predictive analytics. In the present study, the effects of tillage systems on soil properties and crop production and the predictive capabilities of multiple linear regressions (MLR) and artificial neural networks (ANN) are evaluated to estimate organic potato crop yield including soil microbial biomass (MB), soil resistance to penetration, soil organic matter (OM) and tillage system. Potato yield was found to be significantly impacted by tillage and soil properties. The results showed that MLR model estimated crop yield more accuracy than ANN model. Correlation coefficient and root mean squared (RMSE) were 0.97 and 0.077 between the measured and the estimated data by the ANN model, respectively. Generally, the ANN model showed greater potential in determining the relationship between potato yield, tillage and soil properties.