Using artificial neural networks and regression to predict percentage of applied nitrogen leached under turfgrass

Abstract

The objective of this study was to develop an Artificial Neural Network (ANN) model that accurately predicts the percentage of applied nitrogen (N) that leaches through the upper 50 cm of soil under a variety of conditions. The statistical regression models were used for comparison with the ANN model. The Sum of the Squared Error (SSE) between the anticipated values (from research data) and the predicted values (produced by the model) was calculated to be 0.3 for the ANN model and 0.1 for the third order regression. In this particular project, the first and second order regression equations are not useful; however, the third order equation could be used by turf managers along side the ANN model to accurately predict leachate under given field conditions. These models enable the turfgrass manager to determine the effects of management practices on N leaching.