Evaluation of outlier detection method performance in symmetric multivariate distributions


UZABACI E., ERCAN İ., ALPU Ö.

COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, vol.49, no.2, pp.516-531, 2020 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 49 Issue: 2
  • Publication Date: 2020
  • Doi Number: 10.1080/03610918.2018.1487068
  • 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.516-531
  • Bursa Uludag University Affiliated: Yes

Abstract

Determining outliers is more complicated in multivariate data sets than it is in univariate cases. The aim of this study is to evaluate the blocked adaptive computationally efficient outlier nominators (BACON) algorithm, the fast minimum covariance determinant (FAST-MCD) method, and the robust Mahalanobis distance (RM) method in multivariate data sets. For this purpose, outlier detection methods were compared for multivariate normal, Laplace, and Cauchy distributions with different sample sizes and numbers of variables. False-negative and false-positive ratios were used to evaluate the methods' performance. The results of this work indicate that the performance of these methods varies according to the distribution type.