Multivariate pattern analysis for MEG: A comparison of dissimilarity measures

Matthias Guggenmos, Philipp Sterzer, Radoslaw Martin CichyNeuroImage 2018 · OpenReview

Abstract

Highlights•We provide Python and MATLAB tutorials for key analysis steps.•We compared dissimilarity measures and preprocessing choices for MEG MVPA.•Multivariate noise normalisation is a key preprocessing step.•LDA, SVM and WeiRD are recommended classifiers for decoding.•The cross-validated Euclidean distance is a reliable and unbiased choice for RSA.
BibTeX
@inproceedings{guggenmos2018multivariate,
  author = {Matthias Guggenmos and Philipp Sterzer and Radoslaw Martin Cichy},
  title = {{Multivariate pattern analysis for MEG: A comparison of dissimilarity measures}},
  booktitle = {NeuroImage},
  year = {2018},
  url = {https://openreview.net/forum?id=oGqDKtJLK5},
}

Loading related content...

Figure