BIOT: Explaining multidimensional nonlinear MDS embeddings using the Best Interpretable Orthogonal Transformation

Adrien Bibal, Rebecca Marion, Rainer von Sachs, Benoît Frénay

Research output: Contribution to journalArticlepeer-review

Abstract

Dimensionality reduction (DR) is a popular approach to data exploration in which instances in a given dataset are mapped to a lower-dimensional representation or “embedding.” For nonlinear dimensionality reduction (NLDR), the dimensions of the embedding may be difficult to understand. In such cases, it may be useful to learn how the different dimensions relate to a set of external features (i.e., relevant features that were not used for the DR). A variety of methods (e.g., PROFIT and BIR) use external features to explain embeddings generated by NLDR methods with rotation-invariant objective functions, such as multidimensional scaling (MDS). However, these methods are restricted to two-dimensional embeddings. In this paper, we propose BIOT, which makes it possible to explain an MDS embedding with any number of dimensions without requiring visualization.

Original languageEnglish
Pages (from-to)109-118
Number of pages10
JournalNeurocomputing
Volume453
DOIs
Publication statusPublished - 17 Sep 2021

Keywords

  • Explainability
  • Lasso
  • Multidimensional scaling
  • Orthogonal transformations

Fingerprint

Dive into the research topics of 'BIOT: Explaining multidimensional nonlinear MDS embeddings using the Best Interpretable Orthogonal Transformation'. Together they form a unique fingerprint.

Cite this