基于arima模型的北京市蔬菜价格预测分析解析.pdfVIP

基于ARIMA模型的北京市蔬菜价格预测分析 The Forecast of Bejing Vegetable Price Based on ARIMA Model 报告人：沈辰 指导老师：穆月英 大连·辽宁 2010.9.2 研究背景 Research Background  In recent period, the prices of many agricultural products in China keeps going up, especially the vegetables price rised most obviously. Forecast Price of Tomato 3.8 3.3 2.8 2.3 1.8 1.3 0.8 1月 2月 3月 4月 5月 6月 7月 8月 9月 10月 11月 12月 Month 2008 2009 2010 研究背景 Research Background  Question  How to forecast the future price  Method  y=f (x)  Time Series Analysis  ARIMA Model ARIMA Model  ARIMA Model was first advanced by G.E.P. Box and G.M. Jenkins in 1970. Its full name is integrated Autoregressive Moving Average Model.  Model assume that things changes gradually, not suddenly and large-scale. The impact factors in the past, current and future basically not change or changed little. Things change in accordance with the principle of the stability. ARMIA Model  Some symbol:  B: Lag operator(滞后算子) BX X Bk X X t t1 t tk (B)  B  B2 0 1 2 operator polynomial  t :White noise process( 白噪声过程) (1) E ( ) 0