Deakin University
Browse

Bayesian Optimization with Discrete Variables

Download (865.27 kB)
Version 2 2023-10-12, 03:11
Version 1 2023-10-06, 05:00
conference contribution
posted on 2023-10-12, 03:11 authored by P Luong, Sunil GuptaSunil Gupta, Dang NguyenDang Nguyen, Santu RanaSantu Rana, Svetha VenkateshSvetha Venkatesh
Bayesian Optimization (BO) is an efficient method to optimize an expensive black-box function with continuous variables. However, in many cases, the function has only discrete variables as inputs, which cannot be optimized by traditional BO methods. A typical approach to optimize such functions assumes the objective function is on a continuous domain, then applies a normal BO method with a rounding of suggested continuous points to nearest discrete points at the end. This may cause BO to get stuck and repeat pre-existing observations. To overcome this problem, we propose a method (named Discrete-BO) that manipulates the exploration of an acquisition function and the length scale of a covariance function, which are two key components of a BO method, to prevent sampling a pre-existing observation. Our experiments on both synthetic and real-world applications show that the proposed method outperforms state-of-the-art baselines in terms of convergence rate. More importantly, we also show some theoretical analyses to prove the correctness of our method.

History

Volume

11919

Pagination

473-484

Location

Adelaide, South Australia

Open access

  • Yes

Start date

2019-12-02

End date

2019-12-05

ISSN

0302-9743

eISSN

1611-3349

ISBN-13

9783030352875

Language

eng

Publication classification

E1 Full written paper - refereed

Editor/Contributor(s)

Liu J, Bailey J

Title of proceedings

AI 2019 : Advances in Artificial Intelligence : Proceedings of the 32nd Australian Joint Conference

Event

Artificial Intelligence. Joint Conference (2019 : 32nd : Adelaide, South Australia)

Publisher

Springer

Place of publication

Berlin, Germany

Series

Lecture Notes in Computer Science; v.11919

Usage metrics

    Research Publications

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC