A Robust Simulation Optimization Algorithm using Bayesian Method

Please enable javascript in your browser.

Seifi, Farshad | 2019

584 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 52381 (01)
University: Sharif University of Technology
Department: Industrial Engineering
Advisor(s): Akhavan Niaki, Taghi
Abstract:
Huge availability of data in last decade has raised the opportunity to use data for decision making. The idea of using existing data to achieve more coherent reality solution has led to a branch of optimization called data-driven optimization. Presence of uncertain variables makes it crucial to design robust optimization methods for this area. On the other hand, in many real-world problems, the closed-form of the objective function is not available and a meta-model based framework is necessary. Motivated by this, we are using a Gaussian process in a Bayesian optimization framework to design a method that is consistent with the data in predefined confidence level. The goodness of the developed method is that it is computationally tractable in addition to being robust and independent of the objective function’s form. As one of the applications of the proposed algorithm, hyperparameter optimization for deep learning is investigated. The proposed method can help find the optimal hyperparameters that are robust w.r.t data noise
Keywords:
Simulation Optimization ; Data Driven Method ; Black-Box Optimization ; Gaussian process ; Bayesian Optimization ; Hyperparameter Optimization

No TOC