关键词: 模糊时间序列模型, 季节模型, 相对误差, 均方根误差

Abstract: Fuzzy time series model and seasonal model are both based on time series. In order to investigate which one is better in forecasting when time series is periodic, forecasting the passenger flow of a shopping mall in Nanjing uses fuzzy time series model and seasonal model respectively. By comparing the value of relative error and the value of RMSE after computing them which are produced by the two different methods, it finds out that the relative error graphics of seasonal model is smoother than fuzzy time series model’s. And the value of RMSE of seasonal model is also smaller than fuzzy time series model’s. It indicates that it will get a better prediction when the character of data is concerned.

Key words: fuzzy time series model, seasonal model, relative error, Root Mean Square Error（RMSE）