由於一般追?資料模型皆使用大量的資料，非常容易有離群值的發生，這些離群值會影響到模型的估計及效率，因此在本研究中希望能建構出一個快速且有效率的方法，來檢測追?資料模型中的離群值。本研究主要針對追蹤資料模型中離群值的偵測及認定，吾人使用了 spectral Whittle 估計法，根據 Chen (2006) 所發表的文章，將Chang, Tiao, Chen (1998) 的介入模型分析方法應用在追?資料的分析上。在本研究中使用了固定效果模型，而且其誤差的型態可以延申到ARFIMA(p,q)數列的型態。其中建立了概度比的檢定方法，可以快速的降低估計的時間及成本。本研究以Monte Carlo 模擬的方式來驗証，當個別組數及時間長度增加時，估計式的檢定力。在實証的分析上，將以台灣地區主要的九家主機板廠商為分析對象，藉以了解模型的適用情形。 This article provides a procedure for outliers’ detection and identification in the spectral domain where the Whittle maximum likelihood estimator of the panel data model proposed by Chen (2006b) is implemented. We extend Chang, Tiao, and Chen (1988)’s approach here to the spectral domain and through the Whittle approach we can quickly detect and identify the type of outliers. A fixed effects panel data model is used, in which the remainder disturbance is assumed to be an ARFIMA process and the likelihood ratio criterion is obtained directly through the modified inverse Fourier transform. This saves much time, especially when the estimated model implements a huge dataset. Through Monte Carlo experiments the consistency of the estimator will be examined by growing the individual number N and time length T, in which the long memory remainder disturbances are contaminated with two types of outliers: additive outlier (AO) and innovation outlier (IO).
