世帯及び個人属性分布を考慮したPT調査データの拡大係数算出手法の適用

  • 倉内 慎也
    愛媛大学大学院 理工学研究科生産環境工学専攻
  • 萩尾 龍彦
    愛媛大学大学院 理工学研究科生産環境工学専攻
  • 石村 龍則
    愛媛大学大学院 理工学研究科生産環境工学専攻
  • 吉井 稔雄
    愛媛大学大学院 理工学研究科生産環境工学専攻

書誌事項

タイトル別名
  • APPLICABILITY OF THE METHODOLOGY TO MATCH DISTRIBUTIONS OF HOUSEHOLD AND PERSON ATTRIBUTES IN THE GENERATION OF SYNTHETIC POPULATIONS BASED ON PERSON-TRIP SURVEY DATA
  • セタイ オヨビ コジン ゾクセイ ブンプ オ コウリョ シタ PT チョウサ データ ノ カクダイ ケイスウ サンシュツ シュホウ ノ テキヨウ

この論文をさがす

抄録

This study investigated the applicability of iterative proportional updating (IPU) algorithm to calculate the weight for each sample case obtained from the disaggregate survey so as to match distributions of household and person attributes in the generation of synthetic populations. The large-scale person-trip survey data conducted in Matsuyama metropolitan area was empirically applied as the disaggregate data, and the results showed that the excessive categorization of attributes would produce large errors due to the lack of sample case, sometimes inducing the failure to calculate the weight. Although the integration of the attribute categories would alleviate above problems, the excessive integration might induce the loss of original information, that is, heterogeneity of the sample case, undermining the advantage of micro-simulation approaches. Therefore, the consideration of sampling scheme is very crucial for the application of IPU algorithm (e.g. setting the sampling ratio for each attribute categories according to the marginal distributions of the population).

収録刊行物

参考文献 (2)*注記

もっと見る

詳細情報 詳細情報について

問題の指摘

ページトップへ