Short-term load forecasting based on data mining

This paper introduces the multiple linear regression, stepwise linear regression, neural network method, and improves the neural network. Comprehensive analysis of the current prediction methods, the application principle of a detailed analysis and comparison of the various prediction methods advantages and disadvantages. Put forward to improve short-term load forecasting accuracy is not only attach importance to the accumulation of historical data, more should pay attention to choose the right forecasting model, integrated prediction model is the future development direction of power load forecasting method.