Causal Discovery from Conditionally Stationary Time Series
This addresses the problem of inferring causal relationships from complex, nonstationary time series for AI systems, representing an incremental advance over traditional methods limited to stationary data.
The paper tackled causal discovery from nonstationary time series by modeling them as conditionally stationary with latent state variables, resulting in the State-Dependent Causal Inference (SDCI) method that outperformed baseline methods in experiments on particle interaction data and gene regulatory networks.
Causal discovery, i.e., inferring underlying causal relationships from observational data, is highly challenging for AI systems. In a time series modeling context, traditional causal discovery methods mainly consider constrained scenarios with fully observed variables and/or data from stationary time-series. We develop a causal discovery approach to handle a wide class of nonstationary time series that are conditionally stationary, where the nonstationary behaviour is modeled as stationarity conditioned on a set of latent state variables. Named State-Dependent Causal Inference (SDCI), our approach is able to recover the underlying causal dependencies, with provable identifiablity for the state-dependent causal structures. Empirical experiments on nonlinear particle interaction data and gene regulatory networks demonstrate SDCI's superior performance over baseline causal discovery methods. Improved results over non-causal RNNs on modeling NBA player movements demonstrate the potential of our method and motivate the use of causality-driven methods for forecasting.