Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces Rate-Distortion MDL (RDMDL), a new bivariate causal discovery method that improves upon existing MDL-based approaches by refining the estimation of the cause variable's description length. RDMDL leverages rate-distortion theory to measure the description length of the cause, calculating the minimum rate needed to achieve a distortion level representative of the underlying distribution. Experiments on the Tübingen dataset demonstrate that RDMDL achieves competitive performance compared to state-of-the-art methods.
Current MDL-based causal discovery methods may be biased towards the complexity of the causal mechanism, but RDMDL offers a fix by more accurately estimating the description length of the cause variable.
Approaches to bivariate causal discovery based on the minimum description length (MDL) principle approximate the (uncomputable) Kolmogorov complexity of the models in each causal direction, selecting the one with the lower total complexity. The premise is that nature's mechanisms are simpler in their true causal order. Inherently, the description length (complexity) in each direction includes the description of the cause variable and that of the causal mechanism. In this work, we argue that current state-of-the-art MDL-based methods do not correctly address the problem of estimating the description length of the cause variable, effectively leaving the decision to the description length of the causal mechanism. Based on rate-distortion theory, we propose a new way to measure the description length of the cause, corresponding to the minimum rate required to achieve a distortion level representative of the underlying distribution. This distortion level is deduced using rules from histogram-based density estimation, while the rate is computed using the related concept of information dimension, based on an asymptotic approximation. Combining it with a traditional approach for the causal mechanism, we introduce a new bivariate causal discovery method, termed rate-distortion MDL (RDMDL). We show experimentally that RDMDL achieves competitive performance on the Tübingen dataset. All the code and experiments are publicly available at github.com/tiagobrogueira/Causal-Discovery-In-Exchangeable-Data.