aiqbench analysis
This paper proposes a decomposition-based framework to address causal discovery challenges in non-stationary, autocorrelated time series data common in finance and climate science. By separating trends, seasonality, and residuals for targeted analysis, the method effectively isolates genuine causal effects while reducing spurious associations found by standard baselines. The approach represents a significant methodological advance for machine learning models requiring robust temporal inference under complex real-world conditions.
DCD: Decomposition-based Causal Discovery from Autocorrelated and Non-Stationary Temporal Data
arXiv:2602.01433v3 Announce Type: replace-cross Abstract: Multivariate time series in domains such as finance, climate science, and healthcare often exhibit long-term trends, seasonal patterns, and short-term fluctuations, complicating causal inference under non-stationarity and autocorrelation. Existing causal discovery methods typically operate on raw observations, making them vulnerable to spurious edges and misattributed temporal dependencies. We introduce a decomposition-based causal discovery framework that separates each time series into trend, seasonal, and residual components and performs component-specific causal analysis. Trend components are assessed using stationarity tests, seasonal components using kernel-based dependence measures, and residual components using constraint-based causal discovery. The resulting component-level graphs are integrated into a unified multi-scale causal structure. This approach isolates long- and short-range causal effects, reduces spurious associations, and improves interpretability. Across extensive synthetic benchmarks and real-world climate data, our framework more accurately recovers ground-truth causal structure than state-of-the-art baselines, particularly under strong non-stationarity and temporal autocorrelation.
Abstract:Multivariate time series in domains such as finance, climate science, and healthcare often exhibit long-term trends, seasonal patterns, and short-term fluctuations, complicating causal inference under non-stationarity and autocorrelation. Existing causal discovery methods typically operate on raw observations, making them vulnerable to spurious edges and misattributed temporal dependencies. We introduce a decomposition-based causal discovery framework that separates each time series into trend, seasonal, and residual components and performs component-specific causal analysis. Trend components are assessed using stationarity tests, seasonal components using kernel-based dependence measures, and residual components using constraint-based causal discovery. The resulting component-level graphs are integrated into a unified multi-scale causal structure. This approach isolates long- and short-range causal effects, reduces spurious associations, and improves interpretability. Across extensive synthetic benchmarks and real-world climate data, our framework more accurately recovers ground-truth causal structure than state-of-the-art baselines, particularly under strong non-stationarity and temporal autocorrelation.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
arXiv:2602.01433 [cs.LG]
(or arXiv:2602.01433v3 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2602.01433 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 240 words
Related coverage
7 Steps to Mastering Time Series Analysis with Python
KDnuggets · Jun 4, 2026
PIRAMID: Progress and Plans
LessWrong · Jul 27, 2026
[AINews] not much happened today
Latent Space · Sep 18, 2026
LMT: A Bayesian Framework for Causal Discovery from Textual Alarm Records in Manufacturing Systems
arXiv · Jun 10, 2026