dc.contributor | 國立政治大學統計學系 | en_US |
dc.contributor | 行政院國家科學委員會 | en_US |
dc.creator (作者) | 黃子銘 | zh_TW |
dc.date (日期) | 2010 | en_US |
dc.date.accessioned | 30-Aug-2012 09:58:33 (UTC+8) | - |
dc.date.available | 30-Aug-2012 09:58:33 (UTC+8) | - |
dc.date.issued (上傳時間) | 30-Aug-2012 09:58:33 (UTC+8) | - |
dc.identifier.uri (URI) | http://nccur.lib.nccu.edu.tw/handle/140.119/53374 | - |
dc.description.abstract (摘要) | 條件獨立檢定的應用之一為檢定 Granger causality, 這時資料通常並非獨立。在國科會計畫 NSC-95-2119-M-004-001 及 NSC-97-2118-M-004-001 中,提出根據 maximal conditional correlation 建立條件獨立檢定,也在資料為獨立同分布的假設下,推導出檢定統計量的漸進性質。本計畫的目標即為將上述結果推廣到資料為 weakly dependent 的情形,使得這個條件獨立檢定能適用於 Granger causality 的檢定問題。 | - |
dc.description.abstract (摘要) | One use of conditional independence test is to test Granger causality for time series data. In two former NSC projects on testing conditional independence (NSC-95-2119-M-004-001 and NSC-97-2118-M-004-001), a conditional independence test based on maximal conditional correlation is proposed and its asymptotic properties are derived assuming the data series is an IID series. The goal of this project is to show that these asymptotic properties of this test still hold when the data series is weakly dependent, so that the test can be used for time series data. | - |
dc.language.iso | en_US | - |
dc.relation (關聯) | 基礎研究 | en_US |
dc.relation (關聯) | 學術補助 | en_US |
dc.relation (關聯) | 研究期間:9908~ 10007 | en_US |
dc.relation (關聯) | 研究經費:479仟元 | en_US |
dc.subject (關鍵詞) | 條件相關檢定;條件最大相關;相依資料 | en_US |
dc.subject (關鍵詞) | conditional independence test;conditional maximal correlation;dependent data | - |
dc.title (題名) | 相依資料的條件獨立檢定 | zh_TW |
dc.title.alternative (其他題名) | Conditional Independence Test for Dependent Data | en_US |
dc.type (資料類型) | report | en |