Adaptive Expansion Bayesian Optimization for Unbounded Global Optimization
This addresses a practical issue in hyperparameter tuning for machine learning, where setting bounds is non-trivial, by enabling unbounded optimization, though it is incremental as it builds on existing Bayesian optimization frameworks.
The paper tackles the problem of Bayesian optimization requiring fixed variable bounds, which may not include the global optimum, by proposing an approach that adaptively expands the search space while balancing exploration and exploitation. Results show it outperforms or matches state-of-the-art methods on synthetic test functions and an MLP hyperparameter optimization task.
Bayesian optimization is normally performed within fixed variable bounds. In cases like hyperparameter tuning for machine learning algorithms, setting the variable bounds is not trivial. It is hard to guarantee that any fixed bounds will include the true global optimum. We propose a Bayesian optimization approach that only needs to specify an initial search space that does not necessarily include the global optimum, and expands the search space when necessary. However, over-exploration may occur during the search space expansion. Our method can adaptively balance exploration and exploitation in an expanding space. Results on a range of synthetic test functions and an MLP hyperparameter optimization task show that the proposed method out-performs or at least as good as the current state-of-the-art methods.