Causal Compression
This work addresses the challenge of causal inference in time series data, particularly for domains like genomics, but it appears incremental as it builds on existing information-theoretic tools and graph frameworks.
The authors tackled the problem of discovering causal relationships in temporal data by introducing a method based on causal compression, using directed information and a Pearlian graph setting, and demonstrated its application in causal time series segmentation and bipartite graph recovery on gene expression data.
We propose a new method of discovering causal relationships in temporal data based on the notion of causal compression. To this end, we adopt the Pearlian graph setting and the directed information as an information theoretic tool for quantifying causality. We introduce chain rule for directed information and use it to motivate causal sparsity. We show two applications of the proposed method: causal time series segmentation which selects time points capturing the incoming and outgoing causal flow between time points belonging to different signals, and causal bipartite graph recovery. We prove that modelling of causality in the adopted set-up only requires estimating the copula density of the data distribution and thus does not depend on its marginals. We evaluate the method on time resolved gene expression data.