Trend Analysis Of Time Series And Application In Foreign Exchange Transaction

Time series technology is applied in many social fields like economy and medicine. To make corresponding policies, you must master the developing tendency of time series. Therefore, it is significant to make trend analysis of time series technology. The traditional analyses to time series are based on the system, but this paper analyzes time series based on data mining.The time series of foreign exchange rate usually contains many interferential factors, so you must make smooth process to the original data to eliminate the interferential factors. To perform data mining to the time series, you must extract the static attribute which determine the developing tendency of the time series behavior from the time series data. In this paper, the rough set method is adopted to perform data mining to the time series. The data processed by rough set must be the discrete data, so you must disperse the continual data. After the operations you can get a database in which you can perform data mining by rough set method.For the whole information system, you can delete the redundant attribute in the system by attribute reduction. For each policy of the information system, you can delete the redundant attribute value in the system by attribute value reduction. This paper uses one kind of highly effective attribute reduction algorithm for attribute reduction, and then this paper focuses on attribute value reduction and provides one improved rough set value reduction algorithm. This algorithm is developed based on the Heuristic attribute value reduction algorithm and it is approved that its implementing efficiency is enhanced.The effective foreign exchange trade policy is made based on the effective exchange rate forecasting. Therefore, we develop a foreign exchange forecasting and decision-making support system which consists of two subsystems: one is rough set-based foreign exchange forecasting system and the other is foreign exchange decision-making support system. This system can generate accurate short-term exchange rate forecasting result and provides foreign exchange transaction and investment policy for investors. It appraises this system and confirms the superiority of the system through the indexes of the parameter and non-parameter.