Agriculture irrigation water demand forecasting based on rough set theory and weighted LS-SVM

Forecasting agriculture water demand is significant to optimize confirmation of water resources. In this study, we introduce a hybrid model which combines rough set theory and least square support vector machine to forecast the agriculture irrigation water demand. Through a certain district agriculture irrigation water demand dataset experiment, we have proved that the reduction feature set exacted by RST has good subject-independence and intrinsic good separability. Weighted LS-SVM predictor demonstrated promising prediction accuracy, better generalization ability and more rapid execution speed than most of the all benchmarking methods listed in this study.