基于bp神经网络的外汇自动效易系统的设计与实现-计算机技术专业论文.docxVIP

Abstract Abstract With the explosion of modern science and technology development, the economic integration had become a basic feature in the world of contemporary social. Governments, banks, or individuals had to face to the foreign exchange transactions. Any institution or individual, who concerned to the exchange rate, improved their own level of risk management, acquire foreign exchange hedge instruments, and strengthen its exchange rate to adapt to the new situation. In order to overcome the weaknesses of human nature, a lot of number of traders developed a variety of mechanical trading system, then expected to achieve certainty trading profit objective. So, this article was designed and implemented the Forex Automated Trading System. Firstly, the paper studies and system-related technologies; Secondly, by analyzed the normal process of forex trading, it finished the requirements of performance and data, completed the system architecture; finally it researched the neural network. In order to avoid code complexity, this paper used three-tier system architecture that based on client/server; Divided the automated trading systems to improve the efficiency with data collection module, network training module and automated trading module; For the benefit of improving the safety, the paper designed some kernel codes for a higher safety. This thesis completed an automatic trading system with neural network. The system showed the convenience of AI, and brought the mechanical trade and foreign exchange technical analysis into a new level. Forex trading covers a wide range.This paper is only finished the automated trading system design and implementation with neural network . The robustness of the system will be the most important issues of the further research and development. Keywords：Automated Trading Forex Neur al Network Three-tier Architecture 西安电子科技大学 学位论文独创性（或创新性）声明 秉承学校严谨的学分和优良的科学道德，本人声明所呈交的论文是我个人在 导师指导下进行的研究工作及取得的研究成果。尽我所知，除了文中特别加以标 注和致谢中所罗列的内容以外，论文中不包含其它人已经发表