An Entropy Search Portfolio for Bayesian Optimization
This work addresses the challenge of improving optimization efficiency in black-box settings, such as geostatistics and control tasks, with a novel portfolio approach that is robust to poor components.
The paper tackles the problem of selecting acquisition functions in Bayesian optimization by introducing the Entropy Search Portfolio (ESP), which outperforms existing portfolio methods on real and synthetic problems, often matching or exceeding the best unknown acquisition function.
Bayesian optimization is a sample-efficient method for black-box global optimization. How- ever, the performance of a Bayesian optimization method very much depends on its exploration strategy, i.e. the choice of acquisition function, and it is not clear a priori which choice will result in superior performance. While portfolio methods provide an effective, principled way of combining a collection of acquisition functions, they are often based on measures of past performance which can be misleading. To address this issue, we introduce the Entropy Search Portfolio (ESP): a novel approach to portfolio construction which is motivated by information theoretic considerations. We show that ESP outperforms existing portfolio methods on several real and synthetic problems, including geostatistical datasets and simulated control tasks. We not only show that ESP is able to offer performance as good as the best, but unknown, acquisition function, but surprisingly it often gives better performance. Finally, over a wide range of conditions we find that ESP is robust to the inclusion of poor acquisition functions.