Regularized Multivariate Analysis of Variance

J. Engel, Kas J. Houthuijs, Vasilis Vasiliou, Georgia Charkoftaki

Research output: Chapter in Book/Report/Conference proceedingChapterAcademicpeer-review

Abstract

Experimental designs in modern experiments in the life sciences often comprise multiple factors. Usually, the number of response variables vastly exceeds the sample size and well-established methods for analysis of multifactor data such as multivariate analysis of variance (MANOVA) cannot be applied. This issue can be circumvented by combining MANOVA with a shrinkage estimator of the correlation matrix. This chapter discusses the principles of this regularized MANOVA (rMANOVA) approach. Two examples from untargeted metabolomics studies are used to demonstrate application of rMANOVA in practice. Focus lies on identification of significant main and interaction effects and the variables associated to them
Original languageEnglish
Title of host publicationComprehensive Chemometrics
Subtitle of host publicationChemical and Biochemical Data Analysis
EditorsSteven Brown, Romà Tauler, Beata Walczak
PublisherElsevier
Chapter1.18
Pages479-494
Volume1
Edition2nd
ISBN (Print)9780444641663
DOIs
Publication statusPublished - 2 Jun 2020

Fingerprint Dive into the research topics of 'Regularized Multivariate Analysis of Variance'. Together they form a unique fingerprint.

  • Cite this

    Engel, J., Houthuijs, K. J., Vasiliou, V., & Charkoftaki, G. (2020). Regularized Multivariate Analysis of Variance. In S. Brown, R. Tauler, & B. Walczak (Eds.), Comprehensive Chemometrics: Chemical and Biochemical Data Analysis (2nd ed., Vol. 1, pp. 479-494). Elsevier. https://doi.org/10.1016/B978-0-12-409547-2.14577-9