Recursive Causal Discovery with Julia

Explore causal discovery concepts and implementation in Julia through this 12-minute conference talk from JuliaCon 2024. Delve into the fundamentals of determining causal relationships between variables, moving beyond traditional correlation-based statistical methods to understand direct cause-and-effect relationships. Learn about the RecursiveCausalDiscovery.jl package, which implements the Recursive Structural Learning (RSL) algorithm to efficiently discover causal relationships using side-information about underlying graph structures. Examine how this approach improves upon existing methods in the CausalInference.jl package by reducing the number of required conditional independence tests and computational complexity. Follow along as the presentation covers essential background on causality, an overview of causal discovery and the PC algorithm, the principles of recursive causal discovery, and a detailed look at the package implementation. Access accompanying materials including presentation files and source code through the provided GitHub repository to further explore these advanced causal inference techniques in Julia.