頁籤選單縮合
題名 | 季節單根檢定的實務應用= |
---|---|
作者 | 黃春松; |
期刊 | 臺中商專學報 |
出版日期 | 19980600 |
卷期 | 30 1998.06[民87.06] |
頁次 | 頁277-297 |
分類號 | 553.01 |
語文 | chi |
關鍵詞 | 季節單根檢定; 樣本自我相關函數; 辨認程序; Seasonal unit root test; Sample auto-correlation function; Identification procedure; |
中文摘要 | 本文乃是應用HEGY-GLN檢定法,針對與臺灣地區總體經濟有關的實質食品消費、 實質非食品消費、實質國內生產毛額、實質國民生產毛額、實質政府消費、實質醫療及保健 消費等季資料,進行季節單根檢定,並結合 Box-Jenkins 的 ACF 圖形差分次數判斷法相互 參考比較,確定差分的次數,確認時間數列模式形式,建立 ARIMA 模式。 本實務分析同時 獲得三項結果:一、分析最近三十年來臺灣地區總體經濟的幾個重要時間數列季資料的確存 在季節性現象。 二、由分析結果顯示,若以傳統的 Box-Jenkins ACF 差分次數判斷法辨認 模式,可能會出現過度差分或差分不足的現象,三、應用 HEGY-GLN 檢定法時,若選用的顯 著水準不適當,也會造成差分不足的現象。 |
英文摘要 | In this study, we apply HEGY-GLN test to examine whether the time series seasonal data of Taiwan macroeconomics variables exist Seasonal Unit Root, and determine their number of differencing. These variables include real private food consumption, real private nonfood consumption, real gross domestic product, real gross national product, real government consumption expenditure, real medical care and health expense. We use HEGY-GLN test and Box-Jenkins's ACF method simultaneously to specify the typical form for these time series, and construct their ARIMA model. At the same time, We obtain three results:1.Most of the quarterly observations for Taiwan six macroeconomics variables really exist the nature of seasonality. 2.In practice, it is possible to arise over-differencing or under- differencing phenomenon when we use the Jenkins's ACF method to detemine the number of differencing in time series. 3.it is also possible to arise under-differencing phenomenon when we apply the HEGY-GLN test with inadequate significant level to determine the number of differencing in time series. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。