Scalable causal discovery method improves understanding of nonlinear temporal systems
SCDM: Scalable Causal Discovery in Nonlinear Temporal Systems with Meta-Learning
Machine Learning
Summary
Understanding cause and effect in complex systems that change over time is hard, especially when many things interact nonlinearly. The authors introduce SCDM, a new method that learns from all parts of the system at once instead of studying each part separately. This approach makes it easier and faster to find cause-and-effect relationships in large systems. Their method also provides a reliable way to convert learned relationships into clear cause scores, which helps to recover the true system structure.
What this means in practice
- •For healthcare data teams: Identify cause-effect links in high-dimensional brain activity time series to improve understanding of neurological conditions.
- •For financial risk analysts: Analyze large sets of economic indicators over time to discover causal relationships that can inform investment strategies.
Authors
Jingbo Wang, Kegeng Tang, Zihao Wang, Shaogang Ren
Abstract
Causal discovery from nonlinear multivariate time series is challenging in high-dimensional systems, where the number of candidate directed relations grows quadratically with the number of variables. Existing methods are often limited by target-wise model fitting, repeated conditional testing, or expensive graph-search procedures. We propose SCDM, a shared-parameter meta-learning framework for scalable temporal causal discovery. SCDM treats each target variable as a task while learning a shared temporal predictor and a shared causal-strength matrix across tasks. Each meta-episode updates only a subset of target-variable tasks, allowing the model to accumulate structural evidence across the full system without exhaustive target-wise optimization. After training, a post-training graph readout converts the learned predictor into continuous directed causal scores for threshold-independent evaluation. We also provide a finite-error recovery principle showing that, under fully observed delayed-system assumptions and a positive risk-gap condition, thresholding SCDM scores recovers the graph when statistical, meta-optimization, and readout errors are below the population margin. Experiments on controlled synthetic, neuroimaging-inspired, high-dimensional, and real-world temporal benchmarks demonstrate competitive causal recovery and improved scalability in large temporal systems.