Constrained Preferential Bayesian Optimization and Its Application in Banner Ad Design

TL;DR

This paper introduces constrained preferential Bayesian optimization (CPBO), a novel method that incorporates inequality constraints into preference-based optimization, demonstrated through a banner ad design system with positive user study results.

Contribution

The paper presents the first extension of preferential Bayesian optimization to handle inequality constraints, including a new acquisition function and a practical designer-in-the-loop system.

Findings

CPBO effectively finds optimal solutions within feasible regions.

The banner ad design system successfully incorporates subjective preferences and constraints.

User study shows benefits in guiding creative design under constraints.

Abstract

Preferential Bayesian optimization (PBO) is a variant of Bayesian optimization that observes relative preferences (e.g., pairwise comparisons) instead of direct objective values, making it especially suitable for human-in-the-loop scenarios. However, real-world optimization tasks often involve inequality constraints, which existing PBO methods have not yet addressed. To fill this gap, we propose constrained preferential Bayesian optimization (CPBO), an extension of PBO that incorporates inequality constraints for the first time. Specifically, we present a novel acquisition function for this purpose. Our technical evaluation shows that our CPBO method successfully identifies optimal solutions by focusing on exploring feasible regions. As a practical application, we also present a designer-in-the-loop system for banner ad design using CPBO, where the objective is the designer's subjective preference, and the constraint ensures a target predicted click-through rate. We conducted a user study with professional ad designers, demonstrating the potential benefits of our approach in guiding creative design under real-world constraints.