Causal discovery for time series with latent confounders
This work addresses the incremental problem of causal discovery from observational data for researchers in fields like science and AI, where experiments are often infeasible.
The paper evaluated the LPCMCI algorithm for discovering causal relationships in time series with latent confounders, finding it performs much better than random but still far from optimal, with best results on auto-dependencies and struggles on lagged dependencies.
Reconstructing the causal relationships behind the phenomena we observe is a fundamental challenge in all areas of science. Discovering causal relationships through experiments is often infeasible, unethical, or expensive in complex systems. However, increases in computational power allow us to process the ever-growing amount of data that modern science generates, leading to an emerging interest in the causal discovery problem from observational data. This work evaluates the LPCMCI algorithm, which aims to find generators compatible with a multi-dimensional, highly autocorrelated time series while some variables are unobserved. We find that LPCMCI performs much better than a random algorithm mimicking not knowing anything but is still far from optimal detection. Furthermore, LPCMCI performs best on auto-dependencies, then contemporaneous dependencies, and struggles most with lagged dependencies. The source code of this project is available online.