Information-Based Multi-Fidelity Bayesian Optimization
Published in NIPS-17 Workshop on Bayesian Optimization, 2017
Abstract: This paper presents a novel generalization of predictive entropy search (PES) for multi-fidelity Bayesian optimization (BO) called multi-fidelity PES (MF-PES). In contrast to existing multi-fidelity BO algorithms, our proposed MF-PES algorithm can naturally trade off between exploitation vs. exploration over the target and aux- iliary functions with varying fidelities without needing to manually tune any such parameters or input discretization. To achieve this, we first model the unknown target and auxiliary functions jointly as a convolved multi-output Gaussian process (CMOGP) whose convolutional structure is then exploited for deriving an efficient approximation of MF-PES. Empirical evaluation on synthetic and real-world ex- periments shows that MF-PES outperforms the state-of-the-art multi-fidelity BO algorithms.