Can preprocessing EEG signals impact the classification performance?

:question: If you have a question or issue with MNE-Python, please include the following info:

  • MNE version: e.g. 0.24.0
  • operating system: Windows 10

Greetings everyone,

I have read the literature and seen researchers apply various preprocessing techniques before extracting and applying machine learning techniques to perform classification tasks. So, I’m wondering if preprocessing the EEG data is used to improve the classification performance, or it has no effect on the classification performance?


It depends on your data, but generally, some form of preprocessing is recommended (removal of faulty channels, rejection of epochs that show absurdly high voltage peaks, frequency filtering, and potentially removal of eye blinks).

Start with just the basic stuff (bad channels, voltage jumps, frequency filtering) and see if you get any sane results. Then you can start to add more sophisticated preprocessing and check whether decoding performance actually improves.

Best wishes,

Thank you Richard for your valuable comments