Probably Approximately Correct Causal Discovery

TL;DR

This paper introduces the PACC framework, extending PAC learning principles to causal discovery, emphasizing efficiency and providing theoretical guarantees for various causal inference methods under resource constraints.

Contribution

It proposes the PACC framework for resource-efficient causal discovery and offers theoretical guarantees for multiple causal inference methods.

Findings

PACC framework extends PAC principles to causal discovery.

Provides theoretical guarantees for propensity score and instrumental variable methods.

Demonstrates applicability to methods like SCCS with new guarantees.

Abstract

The discovery of causal relationships is a foundational problem in artificial intelligence, statistics, epidemiology, economics, and beyond. While elegant theories exist for accurate causal discovery given infinite data, real-world applications are inherently resource-constrained. Effective methods for inferring causal relationships from observational data must perform well under finite data and time constraints, where "performing well" implies achieving high, though not perfect accuracy. In his seminal paper A Theory of the Learnable, Valiant highlighted the importance of resource constraints in supervised machine learning, introducing the concept of Probably Approximately Correct (PAC) learning as an alternative to exact learning. Inspired by Valiant's work, we propose the Probably Approximately Correct Causal (PACC) Discovery framework, which extends PAC learning principles to the causal field. This framework emphasizes both computational and sample efficiency for established causal methods such as propensity score techniques and instrumental variable approaches. Furthermore, we show that it can also provide theoretical guarantees for other widely used methods, such as the Self-Controlled Case Series (SCCS) method, which had previously lacked such guarantees.