Bayesian Optimization for Cascade-type Multi-stage Processes
Authors:
Shunya Kusakawa,
Shion Takeno,
Yu Inatsu,
Kentaro Kutsukake,
Shogo Iwazaki,
Takashi Nakano,
Toru Ujihara,
Masayuki Karasuyama,
Ichiro Takeuchi
Abstract:
Complex processes in science and engineering are often formulated as multistage decision-making problems. In this paper, we consider a type of multistage decision-making process called a cascade process. A cascade process is a multistage process in which the output of one stage is used as an input for the subsequent stage. When the cost of each stage is expensive, it is difficult to search for the…
▽ More
Complex processes in science and engineering are often formulated as multistage decision-making problems. In this paper, we consider a type of multistage decision-making process called a cascade process. A cascade process is a multistage process in which the output of one stage is used as an input for the subsequent stage. When the cost of each stage is expensive, it is difficult to search for the optimal controllable parameters for each stage exhaustively. To address this problem, we formulate the optimization of the cascade process as an extension of the Bayesian optimization framework and propose two types of acquisition functions based on credible intervals and expected improvement. We investigate the theoretical properties of the proposed acquisition functions and demonstrate their effectiveness through numerical experiments. In addition, we consider an extension called suspension setting in which we are allowed to suspend the cascade process at the middle of the multistage decision-making process that often arises in practical problems. We apply the proposed method in a test problem involving a solar cell simulator, which was the motivation for this study.
△ Less
Submitted 7 March, 2023; v1 submitted 16 November, 2021;
originally announced November 2021.
