Multi-Objective Causal Bayesian Optimization

Shriya Bhatija, Paul-David Zuercher, Jakob Thumm, Thomas Bohné

TL;DR: We incorporate causality into multi-objective Bayesian optimization, enabling decision-making in causal systems with multiple outputs.

Abstract: In decision-making problems, the outcome of an intervention often depends on the causal relationships between system components and is highly costly to evaluate. In such settings, causal Bayesian optimization (CBO) exploits the causal relationships between the system variables and sequentially performs interventions to approach the optimum with minimal data. Extending CBO to the multi-outcome setting, we propose *multi-objective Causal Bayesian optimization* (MO-CBO), a paradigm for identifying Pareto-optimal interventions within a known multi-target causal graph. Our methodology first reduces the search space by discarding sub-optimal interventions based on the structure of the given causal graph. We further show that any MO-CBO problem can be decomposed into several traditional multi-objective optimization tasks. Our proposed MO-CBO algorithm is designed to identify Pareto-optimal interventions by iteratively exploring these underlying tasks, guided by relative hypervolume improvement. Experiments on synthetic and real-world causal graphs demonstrate the superiority of our approach over non-causal multi-objective Bayesian optimization in settings where causal information is available.

Lay Summary: Many real-world decision problems involve choosing actions that influence multiple outcomes, such as improving health while minimizing cost. However, testing different actions can be expensive, especially when outcomes depend on complex cause-and-effect relationships. This work introduces Multi-Objective Causal Bayesian Optimization (MO-CBO), a method that uses known causal information to more efficiently search for the best interventions. MO-CBO rules out suboptimal interventions based on the system’s causal graph, and breaks the problem into simpler optimization tasks. By guiding exploration with a measure of potential improvement, the method finds high-quality solutions with very few data points needed. Experiments show that MO-CBO outperforms standard approaches that do not use causal information, especially when interventions are costly to perform.

Application-Driven Machine Learning: This submission is on Application-Driven Machine Learning.

Link To Code: https://github.com/ShriyaBhatija/MO-CBO

Primary Area: General Machine Learning->Causality

Keywords: Causality, Multi-Objective Bayesian Optimisation

Submission Number: 10596