Presentation Information
[24p-P07-7]Investigation of accelerating catalyst search using Bayesian optimization
〇Junya Kaneda1, Ryoji Asahi1 (1.Nagoya Univ.)
Keywords:
Bayesian optimization,Catalyst
When Bayesian optimization is applied to a wide search range, a problem arises where sufficient experimental acceleration cannot be achieved due to the existence of a large number of local solutions. In this study, we investigated methods to accelerate the search by combining appropriate data preprocessing and optimization algorithms. As a result, it was shown that Bayesian optimization becomes more efficient by adding preprocessing such as clustering the standard data using DBSCAN, and the experiment is accelerated by a factor of 4 compared to a random search.