EEG data is noisy, but decoders still regress usable signal: a win for time-series ML

jonxuxu · x · 2026-11-19

Responding to the EEG reconstruction debate, jonxuxu argues the point isn't whether results are "decent" but that a decoder extracts reliably usable signal from EEG at all — remarkable given its quality. Framed as a whitepill for machine learning on noisy time-series data rather than a claim that EEG is the answer to everything.

Related event: Debate over EEG data quality: can noisy signals yield reliable results(2 posts)→

Original post →

More from Research

Research channel →