Convolutional neural networks for classifying healthy individuals practicing or not practicing meditation according to the EEG data.
Vavilovskii zhurnal genetiki i selektsii December 1, 2023 X Fu, S S Tamozhnikov, A E Saprygin et al. 4 citations
A convolutional neural network trained on event-related brain potentials recorded during a stop-signal paradigm classified people into meditation practitioners and non-practitioners with 82% accuracy. The study developed four non-deep network architectures using data from 100 people (51 meditators, 49 non-meditators) and tested them on an independent sample of 25 people. The best-performing model used a one-dimensional convolutional layer combined with a two-layer fully connected network, though it was prone to overfitting due to limited dataset size. Overfitting was reduced through structural changes, regularization, dropout, and cross-validation. Such models may help assess stress levels and risk for anxiety and depression.