Аннотация:Machine learning methods are starting to be widelyused in the analysis of neuroimaging data. Apart from playing acrucial part in the development of Brain-Computer Interfacetechnologies, machine learning can be also used in academiccontext to link cognitive phenomena to their neurophysiologicalsources. In this study we attempted to use a SVM model to classifyfragments of MEG recording according to the semantic categoriesof the words that were presented to the subject at the moment.The preprocessed data was clustered in spatial and temporaldomains and the clusters were subject to the permutational Ftests. A three-dimensional epochs array was cropped to the timeintervals of significant clusters from the selected channels and hadits dimensionality reduced with Principal Component Analysis(PCA) or Uniform Manifold Approximation and Projection(UMAP). The resulting vector was used to fit the model to solvethe binary classification problem.