| dc.coverage.temporal | 計畫年度：91 起迄日期：20020801～20030731 | en_US |
| dc.creator (作者) | 余清祥 | zh_TW |
| dc.date (日期) | 2002 | en_US |
| dc.date.accessioned | 18-四月-2007 16:36:44 (UTC+8) | en_US |
| dc.date.accessioned | 8-九月-2008 16:06:05 (UTC+8) | - |
| dc.date.available | 18-四月-2007 16:36:44 (UTC+8) | en_US |
| dc.date.available | 8-九月-2008 16:06:05 (UTC+8) | - |
| dc.date.issued (上傳時間) | 18-四月-2007 16:36:44 (UTC+8) | en_US |
| dc.identifier (其他 識別碼) | 912412H004005.pdf | en_US |
| dc.identifier.uri (URI) | http://tair.lib.ntu.edu.tw:8000/123456789/3846 | en_US |
| dc.identifier.uri (URI) | https://nccur.lib.nccu.edu.tw/handle/140.119/3846 | - |
| dc.description (描述) | 核定金額：411000元 | en_US |
| dc.description.abstract (摘要) | 死亡率的變化關係到社會福利與政策的制訂，尤其是死亡率改善幅度快速的台灣地區，死亡率預測的正確性影響至鉅。然而與多數近年經濟突飛猛進的國家類似，台灣地區的死亡率記錄過去並不完整（例如：1950年代最高年齡組為70 歲，1960 年代為80歲，1980 年代為85 歲，1990 年代為100歲）或有瑕疵(例如：1994 年以前嬰兒死亡率的數字)，適用於歐、美、日等已開發國家之死亡率模型不見得適合，由Yue, Hu, and Chang (2001)台灣高齡人口死亡率模型的實證研究可見一斑。本計畫首要目的為比較常見的死亡率模型，包括Lee-Carter、Heligman-Pollard、與關係(Relational)等知名的參數模型，以及縱斷面法與模型生命表等無母數方法，套入台灣地區過去50 年的死亡資料，尋找最符合台灣狀況的0 至100 歲之死亡率模型。實證部分中， 本研究預計使用交叉驗證(Cross-Validation)評定模型的預測優劣；理論部分也將探討如何修正既有的死亡率模型，尋求最佳的估計法，使之能套用至不完整資料。 | - |
| dc.description.abstract (摘要) | Mortality projection plays a critical role in designing social and welfare policies, especially for countries and areas like Taiwan with rapid mortality improvements. However, due to the incomplete data of the elderly mortality and data quality in the earlier days (which are common problems existing in most developing countries), directly applying frequently used methods, such as Lee-Carter and Heligman-Pollard models are highly questionable. See Yue, Hu, and Chang (2001) for a discussion of the mortality rates of elderly in Taiwan. In this study, we will compare some well-known mortality models, including Lee-Carter model, Heligman-Pollard model, Relational model (parametric form), Cohort method, and Model Life Tables (nonparametric form), based on the 1951-2000 Taiwan mortality data, and try to find the method that can describe the mortality characteristics in Taiwan the best. Empirically, we will use cross-validation to judge which method has the best fit. Also, we will seek ways to modify the current methods and improve their estimation in the cases of incomplete data. | - |
| dc.format | applicaiton/pdf | en_US |
| dc.format.extent | bytes | en_US |
| dc.format.extent | 50903 bytes | en_US |
| dc.format.extent | 50903 bytes | - |
| dc.format.extent | 12676 bytes | - |
| dc.format.mimetype | application/pdf | en_US |
| dc.format.mimetype | application/pdf | en_US |
| dc.format.mimetype | application/pdf | - |
| dc.format.mimetype | text/plain | - |
| dc.language | zh-TW | en_US |
| dc.language.iso | zh-TW | en_US |
| dc.publisher (出版社) | 臺北市：國立政治大學統計學系 | en_US |
| dc.rights (權利) | 行政院國家科學委員會 | en_US |
| dc.subject (關鍵詞) | 死亡率;預測;Lee-Carter 模型;Heligman-Pollard 模型;關係模型;縱斷面;模型生命表 | - |
| dc.subject (關鍵詞) | Mortality Rates;Projection;Lee-Carter Model;Heligman-Pollard Model;Cohort;Relational Model;Model Life Tables | - |
| dc.title (題名) | 台灣地區死亡率模型的理論與實證研究 | zh_TW |
| dc.title.alternative (其他題名) | A Theoretical and Empirical Study of Mortlity Models in Taiwan Area | - |
| dc.type (資料類型) | report | en |
