本文详细证明了MA模型的自协方差函数，涉及不同情况下的计算，包括k=0，1≤k≤q和k>q的情况，揭示了自协方差函数与参数θ的关系。
MA模型的自协方差函数证明
∵ E ( ε t ) = 0 ∴ c o v ( X t , X t − k ) = c o v ( ε t − θ 1 ε t − 1 − ⋯ − θ q ε t − q , ε t − k − θ 1 ε t − k − 1 − ⋯ − θ q ε t − k − q ) = c o v ( ε t , ε t − k ) − ∑ j = 1 q θ 0 θ j c o v ( ε t , ε t − k − j ) − ∑ i = 1 q θ i θ 0 c o v ( ε t − i , ε t − k ) + ∑ i = 1 q ∑ j = 1 q θ i θ j c o v ( ε t − i , ε t − k − j ) = E ( ε t , ε t − k ) − ∑ j = 1 q θ 0 θ j E ( ε t , ε t − k − j ) − ∑ i = 1 q θ i θ 0 E ( ε t − i , ε t − k ) + ∑ i = 1 q ∑ j = 1 q θ i θ j E ( ε t − i , ε t − k − j ) \because E(\varepsilon_t)=0\\ \therefore\ cov(X_t,X_{t-k})\\ =cov(\varepsilon_t-\theta_1\varepsilon_{t-1}-\dots-\theta_q\varepsilon_{t-q}, \varepsilon_{t-k}-\theta_1\varepsilon_{t-k-1}-\dots-\theta_q\varepsilon_{t-k-q}) \\=cov(\varepsilon_t,\varepsilon_{t-k})-\displaystyle\sum_{j=1}^q\theta_0\theta_jcov(\varepsilon_{t},\varepsilon_{t-k-j})-\displaystyle\sum_{i=1}^q\theta_i\theta_0cov(\varepsilon_{t-i},\varepsilon_{t-k})+\displaystyle\sum_{i=1}^q\displaystyle\sum_{j=1}^q\theta_i\theta_jcov(\varepsilon_{t-i},\varepsilon_{t-k-j}) \\=E(\varepsilon_t,\varepsilon_{t-k})-\displaystyle\sum_{j=1}^q\theta_0\theta_jE(\varepsilon_{t},\varepsilon_{t-k-j})-\displaystyle\sum_{i=1}^q\theta_i\theta_0E(\varepsilon_{t-i},\varepsilon_{t-k})+\displaystyle\sum_{i=1}^q\displaystyle\sum_{j=1}^q\theta_i\theta_jE(\varepsilon_{t-i},\varepsilon_{t-k-j}) ∵E(εt)=0∴ cov(Xt,Xt−k)=cov(εt−θ1εt−1−⋯−θqεt−q,εt−k−θ1εt−k−1−⋯−θqεt−k−q)=cov(εt,εt−k)−j=1∑qθ0θ
最低0.47元/天 解锁文章
扫一扫
5552
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
