Search papers, labs, and topics across Lattice.
The paper introduces Distributed Dynamic Invariant Causal Prediction in Time-series (DisDy-ICPT) to learn dynamic causal relationships from environmental time series data while addressing spatial confounding. DisDy-ICPT identifies stable causal predictors in a distributed setting without data communication, achieving this through dynamic causal analysis and invariant causal inference. Experiments on synthetic and real-world datasets demonstrate that DisDy-ICPT outperforms baselines in predictive stability and accuracy, suggesting utility in applications like carbon monitoring and weather forecasting.
Discovering stable causal relationships in distributed environmental time series is now possible without direct data sharing, thanks to a new framework that balances dynamic analysis with spatial confounding mitigation.
The extraction of invariant causal relationships from time series data with environmental attributes is critical for robust decision-making in domains such as climate science and environmental monitoring. However, existing methods either emphasize dynamic causal analysis without leveraging environmental contexts or focus on static invariant causal inference, leaving a gap in distributed temporal settings. In this paper, we propose Distributed Dynamic Invariant Causal Prediction in Time-series (DisDy-ICPT), a novel framework that learns dynamic causal relationships over time while mitigating spatial confounding variables without requiring data communication. We theoretically prove that DisDy-ICPT recovers stable causal predictors within a bounded number of communication rounds under standard sampling assumptions. Empirical evaluations on synthetic benchmarks and environment-segmented real-world datasets show that DisDy-ICPT achieves superior predictive stability and accuracy compared to baseline methods A and B. Our approach offers promising applications in carbon monitoring and weather forecasting. Future work will extend DisDy-ICPT to online learning scenarios.
