Yujin Taguchi

Optimization of time-consuming experimental conditions using pseudo-experimental data guided by adaptive polynomial regression

Hirotaka Sugawara, Yujin Taguchi, Kei Minagawa, Yusuke Hiki, Takashi Morikura, Akira Funahashi

Bayesian optimization (BO) is an optimization method that sequentially proposes the next candidate explainable variables for optimizing target variables by balancing exploration and exploitation. BO is often used under a limited evaluation budget, such as hyperparameter tuning of…Read more