Think Global and Act Local: Bayesian Optimisation over High-Dimensional Categorical and Mixed Search Spaces
This work addresses a critical bottleneck in optimization for domains like hyperparameter tuning and materials science, offering a novel approach with empirical gains.
The paper tackles the challenge of high-dimensional black-box optimization in categorical and mixed search spaces by proposing a method that combines local optimization with a tailored kernel design, achieving improved performance and computational efficiency over existing baselines on synthetic and real-world tasks.
High-dimensional black-box optimisation remains an important yet notoriously challenging problem. Despite the success of Bayesian optimisation methods on continuous domains, domains that are categorical, or that mix continuous and categorical variables, remain challenging. We propose a novel solution -- we combine local optimisation with a tailored kernel design, effectively handling high-dimensional categorical and mixed search spaces, whilst retaining sample efficiency. We further derive convergence guarantee for the proposed approach. Finally, we demonstrate empirically that our method outperforms the current baselines on a variety of synthetic and real-world tasks in terms of performance, computational costs, or both.