查詢結果分析
來源資料
相關文獻
- 改良式ARCH模式之應用: 以匯率變動為例
- Fuzzy-genetic Procedure for Change Points Detection
- 臺灣貨幣需求函數之結構轉折點與轉折區間的比較研究:變數模糊時間序列的應用
- Change Points Detection for Nonlinear Time Series and Its Application
- 匯率預測的理論與實務
- 公眾預期心理對臺灣匯率預測之效果評估--ARIMAT模式分析
- 應用基因演算法與倒傳遞類神經網路於匯率預測模型之開發
- Heuristic Methods in Time Series Analysis Processing Using Fuzzy Statistics
- 模型組合與新臺幣匯率預測
- A New Approach in Multiple Attribute Decision Making Using R-norm Entropy and Hamming Distance Measure
頁籤選單縮合
題 名 | 改良式ARCH模式之應用: 以匯率變動為例=On Trimmed ARCH Model: With Application to the Exchange Rate |
---|---|
作 者 | 吳正義; | 書刊名 | 正修學報 |
卷 期 | 11 1998.07[民87.07] |
頁 次 | 頁189-196 |
分類號 | 563.2 |
關鍵詞 | 模糊熵; 匯率預測; ARCH; Trimmed ARCH; Fuzzy entropy; Exchange rate forecasting; |
語 文 | 中文(Chinese) |
中文摘要 | 有關經濟、金融領域的時間數列,常發生模型產生結構性改變;即其動態走勢本 身常會出現很大的不穩定震盪與變異數不確定情形。自Engle(1982)提出自迴歸條件變異 數分析法(AutoRegressive Conditional Heteroscedasticity, ARCH),考慮由於條件變 異數隨時間變動而變動的特性,可以成功地掌握到時間數列趨勢。本文提出改良式自迴歸條 件變異數分析法(Trimmed ARCH),即是找出時間數列之轉折區間,再將轉折區間前段之數 列資料刪去,以做為 ARCH 模型預測的原始資料。如此將較整段未經結構性改變而刪除的數 列資料更具掌握新資訊的優點,所得的預測值也更具準確性。 |
英文摘要 | In analyzing economic or financial time series, we often find its dynamic process exhibits a pattern with structure changed. Recently, researchers are interested in the behavior of nonlinearly which is reflected in the variance. Engel(1982) first proposed the ARCH (AutoRegressive Conditional Heteroscedasticity). He suggested that the variance of a certain time series can be modeled in terms of past observations and the variance of the process is time variant. In this paper, we try to recognize the structure change period by the fuzzy classification; i.e. let the first point and the last point of the structure change period be the cut point, then fit a ARCH model which we called trimmed ARCH model. We use the exchange rate of NT$/US$ as our empirical example. Teh forecasting performance shows that our trimmed ARCH model takes a better prediction result than tradition ARCH model. |
本系統中英文摘要資訊取自各篇刊載內容。