Introduction: I have encephalogram (EEG) data recorded from 32 electrodes. I already did preprocess and epoch extraction. My motivation is segmenting the data for pattern recognition task. I used R-package MLP DNN in MXNet which I am not sure about the quality of the results.

The question: I want to do segmentation(clustering in this case) or classification on this time domain data. For example, the data structure is the matrix with 1200 time samples rows x 32 electrodes columns. If the number of classes is for example 6 (i.e. according to the 10-fold cross-validation method), which Deep learning NNs can be useful and applicable for this task? (could you introduce networks in Matlab or R-packages)

  • 1
    Do you want advice on what kind of neural network to use? Or have you decided what kind of neural network you will use, and you are just looking for a Matlab or R package to achieve that? – Nicolas Raoul Jan 8 '19 at 5:15
  • Thanks for considering! Actually, I need to know which kind of deep NN can be applied in these type of data aim to clustering or classification? (especially Matlab and Rstodio which I am familiar with them) Beside traditional method for clustering such as k-means, Hierarchical clustering, I want to try Deep learning based approach for clustering. For a start point, I have applyed r-package (i.e. MXNet) for labeling. Could you let me know what other DNN I can use for Classification or clustering this type of data? – Rey Jan 8 '19 at 8:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.