Bayesian Optimization
Bayesian Optimization (BO) is a powerful technique for efficiently finding the optimal settings of a complex system or process by sequentially evaluating candidate solutions guided by a probabilistic model. Current research focuses on extending BO's capabilities to high-dimensional spaces, incorporating noise and cost considerations into the optimization process, handling diverse objectives and constraints (including safety), and improving its efficiency through techniques like transfer learning, multi-fidelity approaches, and integration with other algorithms (e.g., genetic algorithms, large language models). BO's impact spans diverse fields, enabling more efficient exploration of design spaces in areas such as materials science, robotics, drug discovery, and machine learning model optimization, ultimately accelerating scientific discovery and technological advancement.
Papers
Bayesian optimized deep ensemble for uncertainty quantification of deep neural networks: a system safety case study on sodium fast reactor thermal stratification modeling
Zaid Abulawi, Rui Hu, Prasanna Balaprakash, Yang Liu
Noise-Aware Bayesian Optimization Approach for Capacity Planning of the Distributed Energy Resources in an Active Distribution Network
Ruizhe Yang, Zhongkai Yi, Ying Xu, Dazhi Yang, Zhenghong Tu
Non-Myopic Multi-Objective Bayesian Optimization
Syrine Belakaria, Alaleh Ahmadianshalchi, Barbara Engelhardt, Stefano Ermon, Janardhan Rao Doppa
Bayesian Optimization of Antibodies Informed by a Generative Model of Evolving Sequences
Alan Nawzad Amin, Nate Gruver, Yilun Kuang, Lily Li, Hunter Elliott, Calvin McCarter, Aniruddh Raghu, Peyton Greenside, Andrew Gordon Wilson
Hyperband-based Bayesian Optimization for Black-box Prompt Selection
Lennart Schneider, Martin Wistuba, Aaron Klein, Jacek Golebiowski, Giovanni Zappella, Felice Antonio Merra
Monte Carlo Tree Search based Space Transfer for Black-box Optimization
Shukuan Wang, Ke Xue, Lei Song, Xiaobin Huang, Chao Qian
Asynchronous Batch Bayesian Optimization with Pipelining Evaluations for Experimental Resource$\unicode{x2013}$constrained Conditions
Yujin Taguchi, Yusuke Shibuya, Yusuke Hiki, Takashi Morikura, Takahiro G. Yamada, Akira Funahashi
Un-evaluated Solutions May Be Valuable in Expensive Optimization
Hao Hao, Xiaoqun Zhang, Aimin Zhou