Publication

Missing values in multi-level simultaneous component analysis

Josse, J., Timmerman, M. E. & Kiers, H. A. L., 15-Nov-2013, In : Chemometrics and Intelligent Laboratory Systems. 129, p. 21-32 12 p.

Research output: Contribution to journalArticleAcademicpeer-review

Copy link to clipboard

Documents

  • Missing values in multi-level simultaneous component analysis

    Final publisher's version, 533 KB, PDF document

    Request copy

DOI

Component analysis of data with missing values is often performed with algorithms of iterative imputation. However, this approach is prone to overfitting problems. As an alternative, Josse et al. (2009) proposed a regularized algorithm in the framework of Principal Component Analysis (PCA). Here we use a similar approach to deal with missing values in multi-level simultaneous component analysis (MLSCA), a method dedicated to explore multivariate multilevel data (e.g., individuals nested within groups). We discuss the properties of the regularized algorithm, the expected behavior under the missing (completely) at random (M(C)AR) mechanisms and possible dysmonotony problems. We explain the importance of separating the deviations due to sampling fluctuations and due to missing data. On the basis of a comparative extensive simulation study, we show that the regularized method generally performs well and clearly outperforms an EM-type of algorithm. (C) 2013 Elsevier B.V. All rights reserved.

Original languageEnglish
Pages (from-to)21-32
Number of pages12
JournalChemometrics and Intelligent Laboratory Systems
Volume129
Publication statusPublished - 15-Nov-2013

    Keywords

  • Multi-set component analysis, Missing data, Regularization, Imputation

View graph of relations

ID: 10735076