GANN-based prediction of fresh water resources

The issue of fresh water resources which has limited development in majority places is one of the most concerned focuses these days. This paper provides a novel approach to designing a specific and rational strategy for prediction of fresh water resources nationwide. The statistics across China are from Official Web sites. By comparing different methods including GM (1, 1), Logistic Regression Model and BP Neural Network, we establish a novel method named GANN which combines the strengths of GM and BP. Besides, WSI (Water Shortage Index) is created to represent the degree of water shortage. Also, experiments of different places are presented in our paper to prove our method.