Large-Scale Targeted Cause Discovery with Data-Driven Learning

2024-08-29

AI Summary

Full-graph causal discovery in large-scale systems suffers from high computational complexity and heterogeneous intervention costs. Method: We propose a target-oriented local causal discovery paradigm that directly identifies the causal parent set of a target variable—bypassing full causal graph reconstruction—and design a linear-time neural network model integrating local inference strategies with causal identifiability modeling, trained under supervision on synthetic data. Contribution/Results: Our approach breaks the bottleneck of traditional global reconstruction, achieving strong out-of-distribution generalization across diverse structural and generative mechanisms. Experiments on *E. coli* and human K562 gene regulatory networks demonstrate significant improvements over full-graph baselines, scalability to thousands of variables, and support for cost-aware, differential intervention optimization. The code is publicly available.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Search and Optimization: Local Search

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

We propose a novel machine learning approach for inferring causal variables of a target variable from observations. Our focus is on directly inferring a set of causal factors without requiring full causal graph reconstruction, which is computationally challenging in large-scale systems. The identified causal set consists of all potential regulators of the target variable under experimental settings, enabling efficient regulation when intervention costs and feasibility vary across variables. To achieve this, we train a neural network using supervised learning on simulated data to infer causality. By employing a local-inference strategy, our approach scales with linear complexity in the number of variables, efficiently scaling up to thousands of variables. Empirical results demonstrate superior performance in identifying causal relationships within large-scale gene regulatory networks, outperforming existing methods that emphasize full-graph discovery. We validate our model's generalization capability across out-of-distribution graph structures and generating mechanisms, including gene regulatory networks of E. coli and the human K562 cell line. Implementation codes are available at https://github.com/snu-mllab/Targeted-Cause-Discovery.

Problem

• Infer causal variables of a target efficiently
• Scale to thousands of variables linearly
• Validate on gene regulatory networks effectively

Innovation

• Neural network infers causality from simulated data
• Local-inference strategy ensures linear scalability
• Targets causal factors without full graph reconstruction