Research on the recommendation of mobile phone tariff package based on time series analysis

With communication developing, the choice of tariff packages has become a topic that everyone concerns today. In this article, by adding an error correction factor to the Åström forecast method, an improved model is proposed to predict the future consumption data for each cell phone user and help them to choose the appropriate package. Simulation experiments done on large number of real data manifest that the improved method can reduce the prediction error to a certain extent and improve the accuracy of multi-step prediction compared with Äström method, and can recommend more appropriate tariff packages for mobile phone users.