| dc.contributor | 統計學系 | |
| dc.creator (作者) | 洪英超 | zh_TW |
| dc.date (日期) | 2012 | |
| dc.date.accessioned | 15-Apr-2016 09:52:36 (UTC+8) | - |
| dc.date.available | 15-Apr-2016 09:52:36 (UTC+8) | - |
| dc.date.issued (上傳時間) | 15-Apr-2016 09:52:36 (UTC+8) | - |
| dc.identifier.uri (URI) | http://nccur.lib.nccu.edu.tw/handle/140.119/84726 | - |
| dc.description.abstract (摘要) | 本計畫將探討”驗證二群多變量時間數列因果關係”中之變數選擇問題。其想法主要是 利用所謂的”向量自我迴歸模型”(Vector Autoregression Model) 將二群多變量時間數 列中的”重要變數”萃取出來，並藉此建構一簡化(且新)的”因果關係。當向量自我迴 歸模型的參數已知時，我們將証明此一變數選擇問題的解可以完全的表達出來。 當向 量自我迴歸模型的參數未知時，我們將介紹一個統計的假設檢定程序來估計(或近似) 此一變數選擇問題的解。 | |
| dc.description.abstract (摘要) | In this project we investigate a variable selection problem in the validation of causal relationship between two groups of multivariate time series data. By utilizing the Vector Autoregression (VAR) model, we introduce how to extract “significant variables” in both groups of time series data so that a “trimmed causal relationship” can be presented based on precedence and predictability. When the parameters of the VAR model are known, we show that explicit conditions for solving this variable selection problem can be obtained; when the parameters are unknown, a statistical hypothesis testing procedure is used to approximate the solution. | |
| dc.format.extent | 3805533 bytes | - |
| dc.format.mimetype | application/pdf | - |
| dc.relation (關聯) | 計畫編號 NSC101-2118-M004-002 | |
| dc.subject (關鍵詞) | Granger領先;向量自我迴歸過程;顯著變數;預測;前進選擇法 | |
| dc.subject (關鍵詞) | Granger causality;Vector Autoregressive Process;significant variables;forecasting;forward selection | |
| dc.title (題名) | 二群多變量時間數列間之簡化因果關係 | zh_TW |
| dc.title.alternative (其他題名) | A Trimmed Causal Relationship between Two Groups of Multivariate Time Series Data | |
| dc.type (資料類型) | report | |
