RCBNB-MB Algorithm Discovers Causal Structures in Non-Stationary Time Series

The RCBNB-MB algorithm identifies latent causal regimes in time series data by relaxing the assumption of a static causal structure. It uses Markov blankets to segment time series into stable regimes, providing increased robustness against discovery errors.