ntienvu / knownoptimum_bo

Release code for ICML2020 Knowing The What But Not The Where in Bayesian Optimization

Knowing The What But Not The Where in Bayesian Optimization

Introduction

Bayesian optimization has demonstrated impressive success in finding the optimum input x∗ and output f∗ = f(x∗) = max f(x) of a black-box function f. In some applications, however, the optimum output f∗ is known in advance and the goal is to find the corresponding optimum input x∗. In this paper, we consider a new setting in BO in which the knowledge of the optimum output f∗ is available. Our goal is to exploit the knowledge about f∗ to search for the input x∗ efficiently. To achieve this goal, we first transform the Gaussian process surrogate using the information about the optimum output. Then, we propose two acquisition functions, called confidence bound minimization and expected regret minimization. We show that our approaches work intuitively and give quantitatively better performance against standard BO methods. We demonstrate real applications in tuning a deep reinforcement learning algorithm on the CartPole problem and XGBoost on Skin Segmentation dataset in which the optimum values are publicly available.

Visualization

demo_visualization_knowing_the_what.....ipynb

Running the algorithms in benchmark functions

demo_on_benchmark_functions.ipynb

Customize your own black-box function

demo_customize_your_own_function.ipynb

Running the comparison using the baselines in benchmark functions

run_all_benchmark_functions.py

After running these scripts to reproduce experiments, the results will be stored as pickles files in "pickle_storage" folder. Then, we can plot all the results using scripts in the "plot" folder.

Dependencies

• numpy >= 1.9.0
• scipy >= 0.14.0
• matplotlib >= 3.1.0
• scikit-learn >= 0.16.1

Paper and Presentation

Visit https://proceedings.icml.cc/static/paper_files/icml/2020/2351-Paper.pdf

Reference

Vu Nguyen and Michael A. Osborne.  "Knowing the what but not the where in Bayesian optimization." International Conference on Machine Learning (ICML), 2020.