Human-Algorithm Collaborative Bayesian Optimization for Engineering Systems

TL;DR

The paper addresses the challenge of incorporating domain expertise into Bayesian optimization for expensive, derivative-free engineering tasks. It introduces a human-in-the-loop framework that couples expert-augmented initial designs with high-throughput, multi-objective Bayesian optimization to generate a diverse set of high-utility alternatives, from which a domain expert selects the next evaluation. The method leverages knee-point selection on a Pareto front to balance information gain and solution diversity, providing interpretable, discrete expert input while preserving BO advantages. Case studies in bioprocess optimization and reactor geometry design show faster convergence and improved accountability, with performance benefits varying by problem dimensionality, noise, and expert alignment.

Abstract

Bayesian optimization has been successfully applied throughout Chemical Engineering for the optimization of functions that are expensive-to-evaluate, or where gradients are not easily obtainable. However, domain experts often possess valuable physical insights that are overlooked in fully automated decision-making approaches, necessitating the inclusion of human input. In this article we re-introduce the human back into the data-driven decision making loop by outlining an approach for collaborative Bayesian optimization. Our methodology exploits the hypothesis that humans are more efficient at making discrete choices rather than continuous ones and enables experts to influence critical early decisions. We apply high-throughput (batch) Bayesian optimization alongside discrete decision theory to enable domain experts to influence the selection of experiments. At every iteration we apply a multi-objective approach that results in a set of alternate solutions that have both high utility and are reasonably distinct. The expert then selects the desired solution for evaluation from this set, allowing for the inclusion of expert knowledge and improving accountability, whilst maintaining the advantages of Bayesian optimization. We demonstrate our approach across a number of applied and numerical case studies including bioprocess optimization and reactor geometry design, demonstrating that even in the case of an uninformed practitioner our algorithm recovers the regret of standard Bayesian optimization. Through the inclusion of continuous expert opinion, our approach enables faster convergence, and improved accountability for Bayesian optimization in engineering systems.

Figures (19)

• Figure 1: Methodology Overview.
• Figure 2: The expert-augmented design of experiments procedure outlined using a 2 dimensional example. Left: the expert defined set of solutions. Center: the initial random set of solutions defined as $\mathbf{X}_{\text{design}}$. Right: the optimal set of solutions comprising of both $\mathbf{X}_{\text{design}}$ alongside $\mathbf{X}_{\text{expert}}$, resulting in the final design of experiments.
• Figure 3: A standard iteration of our algorithm on a one-dimensional case-study.
• Figure 4: A reactor configuration visualized in order to aid the expert in making discrete choices. In doing so, the expert had a reduced need to interpret higher-dimensional decision vectors.
• Figure 5: A reactor configuration visualized in order to aid the expert in making discrete choices. In doing so, the expert had a reduced need to interpret higher-dimensional decision vectors.