Recent advances in Bayesian optimization

X Wang, Y **, S Schmitt, M Olhofer - ACM Computing Surveys, 2023 - dl.acm.org
Bayesian optimization has emerged at the forefront of expensive black-box optimization due
to its data efficiency. Recent years have witnessed a proliferation of studies on the …

Time-varying convex optimization: Time-structured algorithms and applications

A Simonetto, E Dall'Anese, S Paternain… - Proceedings of the …, 2020 - ieeexplore.ieee.org
Optimization underpins many of the challenges that science and technology face on a daily
basis. Recent years have witnessed a major shift from traditional optimization paradigms …

On kernelized multi-armed bandits

SR Chowdhury, A Gopalan - International Conference on …, 2017 - proceedings.mlr.press
We consider the stochastic bandit problem with a continuous set of arms, with the expected
reward function over the arms assumed to be fixed but unknown. We provide two new …

Adversarially robust optimization with Gaussian processes

I Bogunovic, J Scarlett, S Jegelka… - Advances in neural …, 2018 - proceedings.neurips.cc
In this paper, we consider the problem of Gaussian process (GP) optimization with an added
robustness requirement: The returned point may be perturbed by an adversary, and we …

Distributionally robust Bayesian optimization

J Kirschner, I Bogunovic, S Jegelka… - International …, 2020 - proceedings.mlr.press
Robustness to distributional shift is one of the key challenges of contemporary machine
learning. Attaining such robustness is the goal of distributionally robust optimization, which …

Provably efficient online hyperparameter optimization with population-based bandits

J Parker-Holder, V Nguyen… - Advances in neural …, 2020 - proceedings.neurips.cc
Many of the recent triumphs in machine learning are dependent on well-tuned
hyperparameters. This is particularly prominent in reinforcement learning (RL) where a small …

Non-stationary bandits with auto-regressive temporal dependency

Q Chen, N Golrezaei… - Advances in Neural …, 2023 - proceedings.neurips.cc
Traditional multi-armed bandit (MAB) frameworks, predominantly examined under stochastic
or adversarial settings, often overlook the temporal dynamics inherent in many real-world …

Bayesian generational population-based training

X Wan, C Lu, J Parker-Holder, PJ Ball… - International …, 2022 - proceedings.mlr.press
Reinforcement learning (RL) offers the potential for training generally capable agents that
can interact autonomously in the real world. However, one key limitation is the brittleness of …

Safe and efficient model-free adaptive control via Bayesian optimization

C König, M Turchetta, J Lygeros… - … on Robotics and …, 2021 - ieeexplore.ieee.org
Adaptive control approaches yield high-performance controllers when a precise system
model or suitable parametrizations of the controller are available. Existing data-driven …

Corruption-tolerant gaussian process bandit optimization

I Bogunovic, A Krause… - … Conference on Artificial …, 2020 - proceedings.mlr.press
We consider the problem of optimizing an unknown (typically non-convex) function with a
bounded norm in some Reproducing Kernel Hilbert Space (RKHS), based on noisy bandit …