A robust test statistic for independence in high dimensional data

Bulut H.

COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, vol.53, no.2, pp.702-713, 2024 (SCI-Expanded, Scopus) identifier identifier

  • Publication Type: Article / Article
  • Volume: 53 Issue: 2
  • Publication Date: 2024
  • Doi Number: 10.1080/03610918.2022.2028835
  • Journal Name: COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Applied Science & Technology Source, Business Source Elite, Business Source Premier, CAB Abstracts, Compendex, Computer & Applied Sciences, Veterinary Science Database, zbMATH, Civil Engineering Abstracts
  • Page Numbers: pp.702-713
  • Ondokuz Mayıs University Affiliated: No

Abstract

The likelihood ratio test used to test complete independence cannot be used in high-dimensional data. In literature, many test statistics were proposed to cope with this problem. However, these statistics are sensitive to outliers. In this study, we suggest a test statistic that is not sensitive to outliers and can be used in high-dimensional data to test complete independence. We investigate the performance of our suggested statistic in terms of the empirical size and the empirical power by using three simulation studies and a real example dataset. And, we construct an R package to implement our proposed test statistic on real high-dimensional data in applications.