• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to project page

2023 Fiscal Year Final Research Report

Development of new high-dimensional statistical analysis to deal with skewness of sample distribution

Research Project

  • PDF
Project/Area Number 20K11712
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Review Section Basic Section 60030:Statistical science-related
Research InstitutionKanagawa University

Principal Investigator

Hyodo Masashi  神奈川大学, 経済学部, 教授 (00711764)

Project Period (FY) 2020-04-01 – 2024-03-31
Keywords高次元データ / 正規化変換 / 誤差限界 / 多重比較 / 多変量分散分析 / 歪度 / 漸近正規性 / 一致性
Outline of Final Research Achievements

Many approximate tests based on normal approximation have been proposed for hypothesis testing in high-dimensional statistical analysis. It has been revealed that these tests have sufficient accuracy when the dimension p is very large, such as 1000 to 10000. On the other hand, there is a problem that normal approximation does not work for data with medium dimension p, such as 10 to 500, because the distribution of test statistics is distorted. To address these problems, we proposed a new approximation test method that deals with distribution distortion by applying several analytical methods.

Free Research Field

統計科学

Academic Significance and Societal Importance of the Research Achievements

高次元データにおける近似的な仮説検定の多くは、中心極限定理を利用した漸近的な精度保証を行っている。しかし、漸近理論と有限次元のデータに乖離があるため実用性と説得性に欠ける。そこで、本研究では、エッジワース展開や検定統計量の適切な変換を与えることでより正確な漸近分布を導出する。このようなアプローチは古典的な大標本統計学ではよく用いられるが、高次元データにおいては十分に研究されているとは言えないため、古典的な多変量解析における漸近理論を大幅に発展させる可能性があると期待できる。

URL: 

Published: 2025-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi