@article{oai:shobi-u.repo.nii.ac.jp:00000218,
 author = {華山, 宣胤 and HANAYAMA, Nobutane and 渋谷, 政昭 and SHIBUYA, Masaaki},
 journal = {尚美学園大学芸術情報学部紀要, Bulletin of the Faculty of Informatics for Arts, Shobi University},
 month = {Mar},
 note = {本研究では、年齢・時代別に集計された超高齢者の死亡データを極値理論に基づいて解析し、寿命分布の限界を議論する。連続量に対する極値理論を集計された離散データへ適用するために、乱数を加えて生成される擬似連続データを用いる方法と、断片定数死亡強度モデルを用いる方法が提案される。人口動態統計から得られたデータに基づく推定から、寿命分布の限界は、前者の方法では103-35歳、後者の方法では117-19と推定された。, (Age-period)-tabulated data for the oldest-old deaths are analyzed based on the extreme value theory and the limit value of longevity is discussed. In applying the extreme value theory described in the manner of continuous random variables to the analysis of tabulated data, a procedure using quasi-continuous data obtained by adding random values generated from the uniform distribution, and another procedure using the piecewise constant intensity (PCI) model on the Lexis diagram are proposed. From the results of ML estimation based on actual data obtained from the Population movement statistics, the limit value of longevity is estimated to be 103-39 years by the former procedure and 117-19 years by the later procedure., 3, KJ00004916654, 論文, Article},
 pages = {37--46},
 title = {年齢時代区分別超高齢者死亡データに基づく寿命分布の推定},
 volume = {3},
 year = {2004},
 yomi = {ハナヤマ, ノブタネ and シブヤ, マサアキ}
}