Hybrid Local Causal Discovery

2024-12-27

AI Summary

Existing local causal discovery methods suffer from error propagation (in constraint-based approaches) and local equivalence class ambiguity (in score-based approaches), leading to inaccurate skeleton identification and V-structure orientation. This paper proposes the first two-stage hybrid framework integrating constraint-based and score-based paradigms. In Stage I, an OR-rule-based strategy constructs a robust candidate skeleton, avoiding over-pruning inherent in conventional AND-rule methods. In Stage II, BIC scoring refines the structure, and a novel local structural score comparison mechanism is introduced during orientation to explicitly model and eliminate confounding effects from local Markov equivalence classes. Notably, this work is the first to explicitly avoid equivalence-class-induced ambiguity in local orientation, thereby eliminating both cascading errors and stochastic misorientations. Evaluated on 14 benchmark datasets, the method achieves statistically significant improvements in skeleton accuracy and orientation F1-score over seven state-of-the-art baselines.

Technology Category

• Reasoning under Uncertainty: Causality
• Constraint Satisfaction and Optimization: Other Foundations of Constraint Satisfaction
• Machine Learning: Causal Learning

Application Category

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

Abstract

Local causal discovery aims to learn and distinguish the direct causes and effects of a target variable from observed data. Existing constraint-based local causal discovery methods use AND or OR rules in constructing the local causal skeleton, but using either rule alone is prone to produce cascading errors in the learned local causal skeleton, and thus impacting the inference of local causal relationships. On the other hand, directly applying score-based global causal discovery methods to local causal discovery may randomly return incorrect results due to the existence of local equivalence classes. To address the above issues, we propose a Hybrid Local Causal Discovery algorithm, called HLCD. Specifically, HLCD initially utilizes a constraint-based approach combined with the OR rule to obtain a candidate skeleton and then employs a score-based method to eliminate redundant portions in the candidate skeleton. Furthermore, during the local causal orientation phase, HLCD distinguishes between V-structures and equivalence classes by comparing the local structure scores between the two, thereby avoiding orientation interference caused by local equivalence classes. We conducted extensive experiments with seven state-of-the-art competitors on 14 benchmark Bayesian network datasets, and the experimental results demonstrate that HLCD significantly outperforms existing local causal discovery algorithms.

Problem

• Causal Discovery
• AND or OR Rules
• Score-based Methods

Innovation

• Hybrid Local Causal Discovery
• OR Rules
• Score-based Refinement