基于神经网络的期货套期保值决策支持系统  

Future hedge decision support system based on neural network

作　　者：陈晓红[1] 朱霞[1] 

机构地区：[1]中南大学工商管理学院,长沙410083

出　　处：《管理科学学报》2001年第6期18-23,共6页Journal of Management Sciences in China

基　　金：国家杰出青年基金资助项目 (70 12 50 0 7) .

摘　　要：针对期货套期保值业务的特点 ,分析了期货套期保值决策支持系统的整体逻辑 ,尝试用人工神经网络专家系统预测期货行情走势 ,介绍了如何将期货市场与 BP网络有机结合起来构造适合期价预测的模型 ,并且在实证分析中给出了寻找最优预测效果的方法以及预测结论应用分析 .同时 ,提出了以此分析为基础的 DSS系统结构和功能 。This paper analyses total logic of the future hedge decision support system based on the characteristics of hedge operation. Artificial neural network expert system is adopted to forecast the tendency of future market. The ANN model is set and its structure is carefully designed. In the ANN model as prediction of future price, parameters of the model are dynamically changed in order to search preferable effect of prediction. This model has been applied to the real cases of forecasting of future, and the analysis of application of prediction result is given. At the same time, the structure and function of the DSS are given, which have stated significance for supporting decision of hedger.

关 键 词：神经网络 期货 套期保值 决策支持系统 

分 类 号：F830.9[经济管理—金融学]