Sugarcane Yield Prediction Through Data Mining and Crop Simulation Models

The understanding of the hierarchical importance of the factors which influence sugarcane yield can subsidize its modeling, thus contributing to the optimization of agricultural planning and crop yield estimates. The objectives of this study were to identify and ordinate the main variables that condition sugarcane yield, according to their relative importance, as well as to develop mathematical models for predicting sugarcane yield by using data mining (DM) techniques. For this, three DM techniques were applied in the analyses of databases of several sugar mills in the state of São Paulo, Brazil. Meteorological and crop management variables were analyzed through the following DM techniques: random forest; boosting; and support vector machine, and the resulting models were tested through the comparison with an independent data set. Finally, the predictive performances of these models were compared with the performance of a simple agrometeorological model, applied in the same data set. The results allowed to conclude that, within all the variables assessed, the number of cuts was the most important factor considered by all DM techniques. The comparison between the observed yields and those estimated by the DM models resulted in a root mean square error (RMSE) ranging between 19.70 and 20.03 t ha−1, which was much better than the performance of the Agroecological Zone Model, which presented RMSE ≈ 34 t ha−1.