EEG clustering and data compression by the brain
In this talk, our primary goal is to identify classes of models used by the brain to encode the statistical regularities of the environment. We seek to retrieve statistical regularities of the stimuli sequence from data recorded from the brain. Then, we propose a hierarchical clustering procedure, based on a Kolmogorov-Smirnov-type goodness-of-fit test, to clusters in the same group EEG dataset with the same law, and on different groups with different distributions. We test our method in artificial and real datasets exhibiting promising results.