Open Internet by MindsNet

Generalizability of Machine Learning Models on EEG Brain Signals

Machine learning models for motor imagery classification on EEG brain signals fail to generalize across datasets. This limitation hinders the reliability and applicability of these models in real-world scenarios. The issue arises from flawed evaluations, including subject leakage and weak baselines. There is a need for more robust and generalizable models.

Computing & Technology, Computer Science, Machine Learning