|
The main purpose of this research is to build the wholesale price forecast model of vegetable and fruit in Taiwan. This research intended to build the forecast model which can be updated by program automatically with open online-data. Since the wholesale prices of vegetable and fruit in Taiwan affect retail prices, building a wholesale price forecast model can help retailers and consumers react in advanced. According to literature, this research builds two models: “the wholesale price of vegetable and fruit by district forecast model” and “the overall wholesale volume of vegetable and fruit in Taiwan forecast model” with the consideration of the market reality and availability of online open data. Result of this research shows that both models perform well. Two variables in the wholesale price of vegetable and fruit by district regression model, previous wholesale price and wholesale volume, can explain the wholesale price effectively. Furthermore, the variables used in the overall wholesale volume of vegetable and fruit in Taiwan regression model include previous wholesale volume, previous weighted temperature, previous extreme rain volume, day in the week, and crop month. These variables are selected model by model according to their significance. At the end of this study, suggestions to further research are provided based on the current limitations, including: evaluation of psychological expectations, sub-category selection of targets vegetable and fruit, potential influence of import agriculture products, and the time lag effects of climate variables.
|