DSAIMLMay 24, 2018

Learning and Testing Causal Models with Interventions

arXiv:1805.09697v155 citations
Originality Incremental advance
AI Analysis

This work addresses efficient causal inference for researchers in statistics and machine learning, providing theoretical bounds but is incremental as it builds on existing causal network frameworks.

The paper tackles the problem of testing and learning causal Bayesian networks with interventions, showing that O(log n) interventions and O(n/ε^2) samples per intervention suffice to distinguish networks or learn them efficiently, and proves that adaptivity does not reduce the intervention count.

We consider testing and learning problems on causal Bayesian networks as defined by Pearl (Pearl, 2009). Given a causal Bayesian network $\mathcal{M}$ on a graph with $n$ discrete variables and bounded in-degree and bounded `confounded components', we show that $O(\log n)$ interventions on an unknown causal Bayesian network $\mathcal{X}$ on the same graph, and $\tilde{O}(n/ε^2)$ samples per intervention, suffice to efficiently distinguish whether $\mathcal{X}=\mathcal{M}$ or whether there exists some intervention under which $\mathcal{X}$ and $\mathcal{M}$ are farther than $ε$ in total variation distance. We also obtain sample/time/intervention efficient algorithms for: (i) testing the identity of two unknown causal Bayesian networks on the same graph; and (ii) learning a causal Bayesian network on a given graph. Although our algorithms are non-adaptive, we show that adaptivity does not help in general: $Ω(\log n)$ interventions are necessary for testing the identity of two unknown causal Bayesian networks on the same graph, even adaptively. Our algorithms are enabled by a new subadditivity inequality for the squared Hellinger distance between two causal Bayesian networks.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes