Skip to Main Content
A new machine learning method-wavelet neutral network was introduced and some of its characteristics were discussed. Rough set and WNN are combined to establish a rough set-based data pre-processing wavelet network model. It effectively overcome the wavelet network does not distinguish importance of property of samples and slow defect in a large number of data processing operations. After linearly scaling and rough sets theory, the data that affect the mid-long-term electric power load were trained by the tools of WNN.