The reparameterization trick for acquisition functions

File Description SizeFormat 
1712.00424v1.pdfAccepted version667.47 kBAdobe PDFView/Open
Title: The reparameterization trick for acquisition functions
Authors: Wilson, JT
Moriconi, R
Hutter, F
Deisenroth, MP
Item Type: Working Paper
Abstract: Bayesian optimization is a sample-efficient approach to solving global optimization problems. Along with a surrogate model, this approach relies on theoretically motivated value heuristics (acquisition functions) to guide the search process. Maximizing acquisition functions yields the best performance; unfortunately, this ideal is difficult to achieve since optimizing acquisition functions per se is frequently non-trivial. This statement is especially true in the parallel setting, where acquisition functions are routinely non-convex, high-dimensional, and intractable. Here, we demonstrate how many popular acquisition functions can be formulated as Gaussian integrals amenable to the reparameterization trick and, ensuingly, gradient-based optimization. Further, we use this reparameterized representation to derive an efficient Monte Carlo estimator for the upper confidence bound acquisition function in the context of parallel selection.
Issue Date: 1-Dec-2017
Copyright Statement: © 2017 The Author(s)
Keywords: stat.ML
Notes: Accepted at the NIPS 2017 Workshop on Bayesian Optimization (BayesOpt 2017)
Appears in Collections:Faculty of Engineering

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Creative Commonsx