Chow-Liu trees are sufficient predictive models for reproducing key features of functional networks of periictal EEG time-series

Andreas Steimer, Frédéric Zubler, Kaspar Schindler

Abstract: Highlights•We use simple Chow–Liu trees as predictive models of time-sliced, periictal iEEG data.•Correlation matrices are derived from the Chow–Liu trees by Bayesian inference.•Matrices are sparse, they match large empirical correlations but ignore the small.•Matrix-weighted functional network connectivity matches with empirical connectivity.•Matrix-weighted functional network connectivity shows key features of periictal EEG.