my research work is in the field of machine learning classification of electroencephalogram (EEG) signals. I have used on-line EEG signal datasets available in two forms. Here, forms means channels. In one dataset, signals are from single channel Bonn data so I refer it as univariate, another from two channels i.e. bivariate Bern data. The third dataset that I have collected from a neurologies is 16-channel so it become a multivariate signal. I extracted features (various entropies) and supplied it to a random forest classifier. I got varied results for all three dataset. The third being recorded from hospital, the signals were not as standard as the first two on-line datasets. My question is can I do such type of experimentation and analysis i.e. using a single machine learning algorithm on three different datasets and compare results. I had submitted my research paper in a reputed journal and I got "reject" decision with a reason that "cross validation is unclear as the method uses different data". I appreciate your guidance for the question. Thanks

  • $\begingroup$ I used to do some work in similar areas, so I sort of get what you are saying, but it's been a long time so I don't know the answer. I think you need to talk to a specialist in this sort of statistics. $\endgroup$ – Peter Flom Apr 25 '18 at 11:57

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.