基于ARIMA模型的外汇汇率时间序列预测研究

利用数据挖掘技术分析外汇汇率时间序列，从时间序列中获得正确的、隐含的、潜在的信息对于金融领域研究具有重要的现实意义。通过数据挖掘中的ARIMA模型，以某银行的外汇汇率时间序列为研究对象，采用差分方法和建模规则，对外汇的卖出价进行了建模与预测。通过与逐步自回归预测模型相比较，ARIMA模型对外汇汇率时间序列数据具有很强的预测能力。

Prediction of Foreign Exchange Rate Time Series Based on ARIMA Model

Data mining are used to analyze the foreign exchange rate time series and acquire the correct, implicated and hidden information, which has practical significance in the financial field. In this paper, the AR/MA model, which is one of the data mining technologies, is used to study the foreign exchange rate time series of certain bank. In this mod- el, difference and modeling rules are adopted to model and forecast the selling rate. The results show that ARIMA model is capable of predicting accurately data of foreign exchange rate time series.