MVPAlab is a MATLAB toolbox for decoding analyses of multidimensional electroencephalography (EEG) data based on machine learning. It implements multivariate pattern analysis (MVPA) and multivariate cross-classification, and it comes with a graphical user interface for users with little coding experience, as well as high- and low-level routines for those who want to build their own projects.
Features
- Multivariate pattern analysis (MVPA) and multivariate cross-classification.
- Temporal generalization matrices.
- Feature and frequency contribution analyses.
- Configurable analysis: trial averaging, balanced datasets, normalization, smoothing, channel selection, dimensionality reduction, classification models, cross-validation, performance metrics and parallel computation.
- Cluster-based permutation testing.
- Graphical user interface and tools to plot the results.
- Demo scripts and a sample EEG dataset.
MVPAlab is released as open source under the GNU General Public License v3.0.
Reference
D. López-García, J. M. G. Peñalver, J. M. Górriz and M. Ruz. MVPAlab: A machine learning decoding toolbox for multidimensional electroencephalography data. Computer Methods and Programs in Biomedicine, 214, 106549 (2022). DOI: 10.1016/j.cmpb.2021.106549