基于 ARIMA 模型的时间序列数据挖掘方法改进

摘要

ARIMA模型是一种很重要的时间序列数据挖掘模型，但是这个模型往往只是对某个时间点进行研究。事实上一段时间往往影响未来的预测结果，就ARIMA模型的数据挖掘方法进行改进，并用美国IT界的股票价格数据对改进的模型进行了实证分析。结果显示改进后的模型与未来股票价格的预测更加准确。%ARIMA model is a very important time series data mining model.But this model often aims at point in time.In fact, the time often affect future predictions.In this paper, the data mining method of ARIMA model was improved, and was analyzed by stock price da-ta for U.S.IT.The results showed that the improved model and predict future stock prices were more accurate.