RECLAIM: Cyclic Causal Discovery Amid Measurement Noise

2026-03-20

AI Summary

This work addresses the limitations of traditional causal discovery methods, which often fail in real-world scenarios due to their reliance on acyclicity assumptions and noise-free observations. To overcome these challenges, the authors propose RECLAIM, a novel framework that simultaneously models cyclic causal structures and measurement noise for the first time. RECLAIM employs an expectation–maximization (EM) algorithm to jointly infer cyclic causal graphs and latent variables, while leveraging residual normalizing flows to maximize the likelihood of observed data. Under Gaussian additive noise and linear measurement systems, the method enjoys theoretical consistency guarantees. Empirical evaluations demonstrate that RECLAIM significantly outperforms existing approaches on both synthetic benchmarks and real-world protein signaling pathway data.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Uncovering causal relationships is a fundamental problem across science and engineering. However, most existing causal discovery methods assume acyclicity and direct access to the system variables -- assumptions that fail to hold in many real-world settings. For instance, in genomics, cyclic regulatory networks are common, and measurements are often corrupted by instrumental noise. To address these challenges, we propose RECLAIM, a causal discovery framework that natively handles both cycles and measurement noise. RECLAIM learns the causal graph structure by maximizing the likelihood of the observed measurements via expectation-maximization (EM), using residual normalizing flows for tractable likelihood computation. We consider two measurement models: (i) Gaussian additive noise, and (ii) a linear measurement system with additive Gaussian noise. We provide theoretical consistency guarantees for both the settings. Experiments on synthetic data and real-world protein signaling datasets demonstrate the efficacy of the proposed method.

Problem

• causal discovery
• cyclic causality
• measurement noise
• causal graph

Innovation

• cyclic causal discovery
• measurement noise
• expectation-maximization
• residual normalizing flows
• causal structure learning