Development of modeling methods using auxiliary variables and its application

2020 Fiscal Year Final Research Report

• Project/Area Number: 17K12650
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Statistical science
• Research Institution: Hiroshima University
• Principal Investigator: Imori Shinpei 広島大学, 先進理工系科学研究科(理), 准教授 (80747345)
• Project Period (FY): 2017-04-01 – 2021-03-31
• Keywords: 補助変数 / 変数選択 / 情報量規準 / 不完全データ解析

Outline of Final Research Achievements

This study attempted to develop methods exploiting auxiliary variables in order to improve modeling accuracy of the primary variables. A selection method of useful auxiliary variables was developed in incomplete data analysis, where latent variables are included. The goodness of a model constructed by auxiliary variables is measured by an information criterion. A relationship between the proposed criterion and previous criteria was shown. This study also proposed a screening method used when the number of auxiliary variables are large, and derived an information criterion to select useful auxiliary variables under covariate shift.

Free Research Field

数理統計学

Academic Significance and Societal Importance of the Research Achievements

情報資源の有効利用の観点から，補助情報の活用は重要な問題である．しかしながら，補助変数の活用は常に主要変数のモデリング精度を向上させるとは限らず，悪影響を与える可能性もあることから，補助変数の適切な活用が求められる．したがって，本研究で行った有用な補助変数の選択手法など，補助変数を活用する手法の開発は大きな意義があると考えられる．また，関連分野でも一定の研究成果を得ており，今後の補助変数の活用に関する研究が発展する手助けになると期待できる．