Learning search spaces for Bayesian optimization: Another view of hyperparameter transfer learning

Abstract

Bayesian optimization (BO) is a successful methodology to optimize black-box functions that are expensive to evaluate. While traditional methods optimize each black-box function in isolation, there has been recent interest in speeding up BO by transferring knowledge across multiple related black-box functions. In this work, we introduce a method to automatically design the BO search space by relying on evaluations of previous black-box functions. We depart from the common practice of defining a set of arbitrary search ranges a priori by considering search space geometries that are learned from historical data. This simple, yet effective strategy can be used to endow many existing BO methods with transfer learning properties. Despite its simplicity, we show that our approach considerably boosts BO by reducing the size of the search space, thus accelerating the optimization of a variety of black-box optimization problems. In particular, the proposed approach combined with random search results in a parameter-free, easy-to-implement, robust hyperparameter optimization strategy. We hope it will constitute a natural baseline for further research attempting to warm-start BO.

Cited by in corpus (13)

• Using a thousand optimization tasks to learn hyperparameter search strategies
• Amazon SageMaker Autopilot: a white box AutoML solution at scale
• An Asymptotically Optimal Multi-Armed Bandit Algorithm and Hyperparameter Optimization
• Optimizing Discrete Spaces via Expensive Evaluations: A Learning to Search Framework
• Practical Transfer Learning for Bayesian Optimization
• HPO-B: A Large-Scale Reproducible Benchmark for Black-Box HPO based on OpenML
• Practical and sample efficient zero-shot HPO
• Bayesian Active Learning by Disagreements: A Geometric Perspective
• Gaussian Process Sampling and Optimization with Approximate Upper and Lower Bounds
• A multi-objective perspective on jointly tuning hardware and hyperparameters
• A resource-efficient method for repeated HPO and NAS problems
• JUMBO: Scalable Multi-task Bayesian Optimization using Offline Data
• Resonance: Replacing Software Constants with Context-Aware Models in Real-time Communication