Meta-Learning Priors for Safe Bayesian Optimization
This work addresses the problem of safe parameter optimization in robotics, which is incremental as it builds on existing meta-learning and safe BO methods.
The paper tackled the challenge of optimizing controller parameters under safety constraints in robotics by meta-learning priors for safe Bayesian optimization from offline data, demonstrating that these priors accelerate convergence while maintaining safety on benchmark functions and a high-precision motion system.
In robotics, optimizing controller parameters under safety constraints is an important challenge. Safe Bayesian optimization (BO) quantifies uncertainty in the objective and constraints to safely guide exploration in such settings. Hand-designing a suitable probabilistic model can be challenging, however. In the presence of unknown safety constraints, it is crucial to choose reliable model hyper-parameters to avoid safety violations. Here, we propose a data-driven approach to this problem by meta-learning priors for safe BO from offline data. We build on a meta-learning algorithm, F-PACOH, capable of providing reliable uncertainty quantification in settings of data scarcity. As core contribution, we develop a novel framework for choosing safety-compliant priors in a data-riven manner via empirical uncertainty metrics and a frontier search algorithm. On benchmark functions and a high-precision motion system, we demonstrate that our meta-learned priors accelerate the convergence of safe BO approaches while maintaining safety.