Based On Shapley Value-combination Forecasting Model Research On Agricultural Product Logistics Demand In Guizhou

In recent years,Guizhou Province has closely focused on promoting the overall layout of the “five in one” and coordinating the “four comprehensive” strategic layout,and the economic development has been rapid.Accompanied by the rapid economic development in Guizhou Province,residents' demand for agricultural products continues to increase,and the demand for agricultural product logistics in Guizhou Province is also increasing.With the strong support of the provincial government,Guizhou Jiatong and logistics infrastructure have been continuously improved,but the agricultural product market still has the problem of imbalance in supply and demand.The logistics of agricultural products in Guizhou province started late and failed to form a certain scale system,and lack of direct data as research support.In order to promote the development of the agricultural product logistics industry in Guizhou Province,and to achieve an effective resource allocation and supply and demand balance in the agricultural product market,this paper analyzes the agricultural product logistics in Guizhou Province.Demand is forecasted.In order to reduce the error caused by the subjective or objective factors in the single forecasting method,this paper selects the data of related factors in recent years and uses Shapley value combined with smoothing index and multiple regression to build a combined forecasting model to forecast the logistics demand of agricultural products in Guizhou Province.And analyze the current status of Guizhou agricultural product logistics development and existing problems,put forward countermeasures and suggestions.In the theoretical part of the thesis,the paper first introduces the theory of agricultural product logistics,logistics demand and forecasting model.Then,the overall status of Guizhou's economy and the status quo of agricultural product logistics development are described and SWOT analysis is conducted.It also introduces and compares the commonly used logistics demand forecasting methods,combines the advantages and disadvantages of various forecasting methods and the currently available data information,and selects the smoothing index and multivariate regression forecasting model for conceptual and model construction and test analysis.In the empirical part of the thesis,the paper takes the agricultural product logistics in Guizhou as the research object and expounds the current status of Guizhou agricultural product logistics.In order to reduce the errors in the single forecasting model,this paper builds a combined forecasting model based on the Shapley value,and uses two kinds of single model and combined forecasting model to combine the actual data of Guizhou's agricultural product logistics demand from 2002 to 2016 to forecast the future agricultural product logistics demand.Based on the forecasting results and development status,we will further propose countermeasures and suggestions for the problems existing in the development of agricultural product logistics in Guizhou.