DCD: Decomposition-based Causal Discovery from Autocorrelated and Non-Stationary Temporal Data
Muhammad Hasan Ferdous [email protected], Md Osman Gani [email protected] · 2026
Read DCD: Decomposition-based Causal Discovery from Autocorrelated and Non-Stationary Temporal Data by Muhammad Hasan Ferdous [email protected], Md Osman Gani [email protected] (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-01433.txt.
Description
Muhammad Hasan Ferdous and co-authors propose a decomposition-based causal framework for multivariate time series. Finance, climate, and healthcare series blend trends, seasonal patterns, and swings, muddying causal inference. The framework splits each series into trend, seasonal, and residual parts for analysis. Trends get stationarity tests, season parts get kernel dependence measures, and residuals get constraint-based discovery. Merged component graphs form a unified multi-scale structure separating long- and short-range effects. Synthetic benchmarks and real climate data show it recovering ground-truth structure better than baselines.
Catalog record
| CREATOR | Muhammad Hasan Ferdous [email protected], Md Osman Gani [email protected] |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 64721 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602.01433