非平稳信号分析与处理,可用于特征提取，将AR模型扩展应用于非平稳时间序列，得到具有时变系数的时变自回归(time-varying autoregressive, TVAR)模型。(nonstationary random signal analysis and processing)

ARIMA整合移动平均自回归模型，时间序列预测分析方法之一，可用于股价预测。(ARIMA integrates moving average autoregressive model and time series forecasting analysis method, which can be used for stock price forecasting.)