Explainable Federated Bayesian Causal Inference and Its Applications in Advanced Manufacturing

Abstract

Causal inference has recently gained notable attention across various fields like biology, healthcare, and environmental science, especially within explainable artificial intelligence (xAI) systems, for uncovering the causal relationships among multiple variables and outcomes. Yet, it has not been fully recognized and deployed in the manufacturing systems. In this paper, we propose an explainable, scalable, and flexible federated Bayesian (xFB) learning framework to explore causality through the global treatment effect for manufacturing system. By leveraging federated Bayesian learning, we efficiently sample global parameters to derive the propensity score for each client. We then personalize the global parameter and estimate the treatment effect using propensity score matching (PSM). Trough simulations with various datasets and a real-world Electrohydrodynamic (EHD) printing data, we demonstrate that our method outperforms basic Bayesian causal inference and several state-of-the-art federated learning benchmarks.