Prediction of Temperature Time Series Based on Wavelet Transform and Support Vector Machine

To predict the time series, a model combining the wavelet transform and support vector machine is set up. First, wavelet transform is applied to decompose the series into sub series with different time scales. Then, the SVM is applied to the sub series to simulate and predict future behavior. And then by the inverse wavelet transform, the series are reconstructed, which is the prediction for the time series. The prediction precision of the new model is higher than that of the SVM model and the artificial neural network model for many processes, such as runoff, precipitation, temperature. And the new Wavelet-SVM model is applied to analyze the month temperature time series of city Tangshan for example. The universal applicability of the new Wavelet-SVM model and the improvement direction are discussed in the end.