Removing eye activity from EEG signals via ICA

In this previous post, I used linear regression to remove ocular artifacts from EEG signals. A popular alternative to this approach is independent component analysis (ICA). Components that represent ocular activity can be identified and eliminated to reconstruct artifact-free EEG signals. This approach is described in more detail in Jung et al. (2000). A comprehensive comparison between the two methods is beyond the scope of this post.