Logistics demand forecast is the foundation of logistics management.Microscopically, the projection of logistics needs is an important basis of the variousdepartments of the enterprise (including logistics, marketing, production and financialdepartment) to plan and control. From the macro perspective, the correct logisticsdemand forecasting is the premise for making the policy of logistics industrydevelopment. Any kind of prediction methods are based on certain assumptions, andany kind of assumptions can not cover the intricate relationships in the real world. Inthe prediction, if we take it for granted that the forecasting method because of theerror compares greatly, this may cause a part of useful information lost. So academiaproposed the concept of, which is combining different prediction models and makingthe best use of information provided by kinds forecasting methods to obtain thecombination forecasting method according to an appropriate weighted average form.This paper offers the solutions of building Yunnan agricultural logistics system,which is applying the combination forecasting model based on Shapley Value WeightRedistribution to the Yunnan foreign trade data concerning in recent years,avoiding asingle prediction method,which is held to be able to provide the inspiration for thedevelopment of Yunnan’s economic and agricultural.First, this paper introduces the background and significance of the topics, thendiscusses the research status and development of the logistics demand forecasting,combination forecasting model and logistics system of Yunnan agricultural products,points out the main content and method of the paper, and finally summarizes theinnovation and deficiencies.In the theoretical part of the paper, the paper defines the concept and characteristicsof the agricultural product logistics, agricultural logistics system. At the same time,comparisons are made to current situation of domestic and foreign agriculturallogistics system. During the analysis of logistics demand forecasting methods andeffect evaluation, the paper compares the existing quantitative analytical method, thenselects regression analysis the gray prediction model and a simple combination forecasting model. Through the comparative analysis, the paper chooses thecombination forecasting model based on Shapley Value Weight Redistribution. Thenthe basic principles of the model are introduced. Taking advantage of the trade data inrecent years, the paper makes an effect evaluation for three single model andcombination forecasting model.The paper’s empirical research selected Yunnan agricultural products logistics asan analytical object, making a brief analysis for the trade status and characteristics, atthe same time, predicting exports of agricultural products in Yunnan based on thetrade data from2007to2011by combination forecasting model. On that basis, thepaper offers the solutions of building Yunnan agricultural logistics system. At the endof the thesis, this study summarizes and points out future research directions.The paper’s innovative achievements are combining modern logistics theory andquantitative data prediction model based on data mathematical analysis and ensuringthe accuracy of the conclusions. In particular, taking Yunnan agricultural productslogistics for example, the paper provides proposals with very good application valuefor the logistics system of agricultural products. |