Robust Causal Discovery in Real-World Time Series with Power-Laws

2025-07-16

AI Summary

Causal discovery in real-world power-law-distributed time series is highly susceptible to noise and suffers from poor robustness. Method: This paper introduces the first frequency-domain-driven robust causal discovery framework that explicitly incorporates the ubiquitous power-law spectral characteristic of time-series data. It models power spectral density to extract power-law features and integrates statistical causal testing with self-organizing dynamical systems theory to effectively suppress spurious causal signals under nonstationary noise. Contribution/Results: Extensive experiments on synthetic benchmarks and real-world multivariate time series—including financial and climate datasets with known ground-truth causal structures—demonstrate that the proposed method significantly outperforms state-of-the-art (SOTA) approaches in both accuracy and stability of causal direction identification. By unifying spectral analysis, statistical inference, and nonlinear dynamics, it establishes a novel paradigm for interpretable causal modeling of complex dynamic systems.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Security and Privacy: Large-scale security measurements

Abstract

Exploring causal relationships in stochastic time series is a challenging yet crucial task with a vast range of applications, including finance, economics, neuroscience, and climate science. Many algorithms for Causal Discovery (CD) have been proposed, but they often exhibit a high sensitivity to noise, resulting in misleading causal inferences when applied to real data. In this paper, we observe that the frequency spectra of typical real-world time series follow a power-law distribution, notably due to an inherent self-organizing behavior. Leveraging this insight, we build a robust CD method based on the extraction of power -law spectral features that amplify genuine causal signals. Our method consistently outperforms state-of-the-art alternatives on both synthetic benchmarks and real-world datasets with known causal structures, demonstrating its robustness and practical relevance.

Problem

• Identifying causal relationships in noisy time series data
• Overcoming sensitivity of existing causal discovery methods to noise
• Exploiting power-law spectral features for robust causal inference

Innovation

• Leverages power-law spectral features
• Amplifies genuine causal signals
• Outperforms state-of-the-art CD methods