Uncertainty-Aware Multi-Agent Causal Discovery for Non-Stationary Time Series: Regime-Adaptive Stability Consensus
DOI:
https://doi.org/10.66421/int.j.datasci.intell.syst76Keywords:
causal discovery, multi-agent consensus, non-stationarity, moving-block bootstrap, lagged graph, stability selectionAbstract
Discovering evolving directed dependencies requires adapting to changing autoregressive mechanisms and quantifying estimator disagreement. Building on the heterogeneous multi-agent consensus and verification architecture of Liang et al., we combine three complementary regression agents (ordinary least squares, ridge and lasso), moving-block stability scores and an automatically estimated piecewise-stationary boundary selected by a penalized multivariate residual criterion. Across 12 simulated five-node, two-regime VAR systems, the automatically segmented uncertainty-aware estimator attains a mean edge F1 of 0.921, compared with 0.468 for its pooled counterpart, and the estimated breakpoint matches the planted location in 11 of 12 systems and lies within five samples in the remaining one. Comparisons with local OLS (F1 0.933) show that regime localization is the main source of the improvement, while the consensus–verification layer adds edge-level stability and sign-consistency evidence. An exploratory U.S. macroeconomic study illustrates how lagged dependence patterns change across historical subsamples, yielding directed predictive-dependence graphs with explicit resampling stability.
References
C. W. J. Granger, “Investigating causal relations by econometric models and cross-spectral methods,” Econometrica, vol. 37, no. 3, pp. 424–438, 1969. https://doi.org/10.2307/1912791
J. Pearl, Causality: Models, Reasoning, and Inference, 2nd ed. Cambridge University Press, 2009. https://doi.org/10.1017/CBO9780511803161
F. Zhou, “Moment-constrained side-force allocation for wind-assisted merchant ships: Hydrodynamic performance of retractable anti-leeway fins and flap rudders,” Innovation and Technology Studies, vol. 3, no. 1, pp. 27–41, 2026.
J. Runge et al., “Detecting and quantifying causal associations in large nonlinear time series datasets,” Science Advances, vol. 5, no. 11, eaau4996, 2019. https://doi.org/10.1126/sciadv.aau4996
R. Pamfil et al., “DYNOTEARS: Structure learning from time-series data,” in Proc. AISTATS, PMLR 108, pp. 1595–1605, 2020. https://proceedings.mlr.press/v108/pamfil20a.html
B. Huang et al., “Causal discovery from heterogeneous/nonstationary data,” J. Mach. Learn. Res., vol. 21, no. 89, pp. 1–53, 2020. https://www.jmlr.org/papers/v21/19-232.html
A. Hyvärinen, K. Zhang, S. Shimizu, and P. O. Hoyer, “Estimation of a structural vector autoregression model using non-Gaussianity,” J. Mach. Learn. Res., vol. 11, pp. 1709–1731, 2010. https://www.jmlr.org/papers/v11/hyvarinen10a.html
N. Meinshausen and P. Bühlmann, “Stability selection,” J. R. Stat. Soc. B, vol. 72, no. 4, pp. 417–473, 2010. https://doi.org/10.1111/j.1467-9868.2010.00740.x
Y. Liang, Y. Jiao, W. Ping, H. Fan, and X. Han, “Adaptive event-driven labeling: A neuro-symbolic multiagent framework for causal inference in non-stationary time series,” IEEE Access, vol. 14, pp. 104468–104482, 2026. https://doi.org/10.1109/ACCESS.2026.3709267
R. Tibshirani, “Regression shrinkage and selection via the lasso,” J. R. Stat. Soc. B, vol. 58, no. 1, pp. 267–288, 1996. https://doi.org/10.1111/j.2517-6161.1996.tb02080.x
A. E. Hoerl and R. W. Kennard, “Ridge regression: Biased estimation for nonorthogonal problems,” Technometrics, vol. 12, no. 1, pp. 55–67, 1970. https://doi.org/10.1080/00401706.1970.10488634
L. Breiman, “Bagging predictors,” Machine Learning, vol. 24, pp. 123–140, 1996. https://doi.org/10.1007/BF00058655
C. Truong, L. Oudre, and N. Vayatis, “Selective review of offline change point detection methods,” Signal Processing, vol. 167, 107299, 2020. https://doi.org/10.1016/j.sigpro.2019.107299
J. M. Mooij, J. Peters, D. Janzing, J. Zscheischler, and B. Schölkopf, “Distinguishing cause from effect using observational data: Methods and benchmarks,” J. Mach. Learn. Res., vol. 17, no. 32, pp. 1–102, 2016. https://www.jmlr.org/papers/v17/14-518.html
J. Peters, P. Bühlmann, and N. Meinshausen, “Causal inference by using invariant prediction: Identification and confidence intervals,” J. R. Stat. Soc. B, vol. 78, no. 5, pp. 947–1012, 2016. https://doi.org/10.1111/rssb.12167
J. Wang, J. Liu, Y. Ge, and B. Jiang, “Risk-aware ensemble learning for credit default prediction under economic uncertainty,” World Journal of Economics and Business Research, vol. 1, no. 1, pp. 18–32, 2023.
G. Sugihara et al., “Detecting causality in complex ecosystems,” Science, vol. 338, no. 6106, pp. 496–500, 2012. https://doi.org/10.1126/science.1227079
S. L. Lauritzen, Graphical Models. Oxford University Press, 1996. https://global.oup.com/academic/product/graphical-models-9780198522195
D. B. Rubin, “Estimating causal effects of treatments in randomized and nonrandomized studies,” J. Educ. Psychol., vol. 66, no. 5, pp. 688–701, 1974. https://doi.org/10.1037/h0037350
L. Kilian, “Not all oil price shocks are alike: Disentangling demand and supply shocks in the crude oil market,” American Economic Review, vol. 99, no. 3, pp. 1053–1069, 2009. https://doi.org/10.1257/aer.99.3.1053
X. Yue and R. Zhu, “An explainable hybrid machine learning framework for cash-flow-conditioned multi-horizon liquidity risk early warning: An SME-oriented benchmark study,” World Journal of Management Science, vol. 2, no. 1, pp. 63–72, 2024.
S. Seabold and J. Perktold, “Statsmodels: Econometric and statistical modeling with Python,” in Proc. 9th Python in Science Conf., pp. 92–96, 2010. https://doi.org/10.25080/Majora-92bf1922-011
Y. Liu, S. Ren, X. Wang, and M. Zhou, “Temporal Logical Attention Network for Log-Based Anomaly Detection in Distributed Systems,” Sensors, vol. 24, no. 24, Art. no. 7949, 2024, doi: 10.3390/s24247949.
L. Qiu, S. Zhang, Y. Yang, and M. Wang, “Graph-theoretic active learning for the closed-loop discovery of stochastic heterogeneous composites,” PLOS ONE, 2026, doi: 10.1371/journal.pone.0353692.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Emily Tran, Seojun Lewis, Emma Zhang, Victor Gupta

This work is licensed under a Creative Commons Attribution 4.0 International License.
All articles are published under the Creative Commons Attribution 4.0 (CC BY 4.0) License. Users may share, copy, adapt, and distribute the work for any purpose, provided proper credit is given to the original author(s) and source. Authors retain copyright.
