These days, the motto for academic researchers should be, “When in doubt, try more Machine Learning!”

We’ve been trying to understand “brain waves” by analyzing EEGs since they were invented circa 1875.  In the last fifty years, we’ve been using computers to look for patterns in EEGs. And, at the same time, EEGs have become cheaper, easier to do, and ever more detailed.  More data!

I should note that I, too, have to done real research (we wanted to combine many kinds of body sensors, including EEGs), as well as one rather successful “joke” project ([1-4]).  I can say that extracting meaning from EEGs is hard.

One thing we tried was various machine learning approaches, looking for patterns in EEG and other sensor traces.  This is hard, too.  There is so much data, and so many ways to do ML.  And, given that EEG traces come from the sum…