Global continuous optimization with error bound and fast convergence

川口 賢司*; 丸山 結; Zheng, X.

Journal of Artificial Intelligence Research, 56, p.153 - 195, 2016/06

https://doi.org/10.1613/jair.4742

This paper considers global optimization with a black-box unknown objective function that can be non-convex and partly non-smooth. Such a difficult optimization problem arises in many real-world applications, such as parameter tuning in machine learning, engineering design problem, and planning with a complex physics simulator. This paper proposes a new global optimization algorithm, called Locally Oriented Global Optimization (LOGO), to achieve both fast convergence in practice and finite-time error bound in theory. The advantage and usage of the new algorithm are illustrated via theoretical analysis and an experiment conducted with 10 bench-mark test functions. Further, we modify the LOGO algorithm to specifically solve a planning problem with continuous state/action space and long time horizon while maintaining its finite-time error bound. We apply the proposed planning method to severe accident management of a nuclear power plant. The result of the application study demonstrates the practical utility of our method.