矩阵型截面数据时间序列自回归模型

Autoregression Model of Time Series with Matrix Cross-Section Data

矩阵型截面数据时间序列的优点在于可以同时刻画多个对象的多个属性.本文重点研究了矩阵型截面数据时间序列的自回归模型,给出了该模型的参数估计、模型识别、白噪声检验三个方面的理论结果.最后再利用矩阵型截面数据时间序列自回归模型,对两支银行股的日收益率序列和日成交量变化率序列进行建模分析.

The advantage of time series with matrix cross-section data is that multiple attributes of multiple objects can be characterized simultaneously.This paper focuses on autoregression model of time series with matrix cross-section data and presents the methods of parameter estimation,model identification and white noise test.Finally,the daily yield series and daily volume change rate series of two bank stocks are analyzed by this model.