Safe optimization developments and applications

In many machine learning problems, an objective is required to be optimized with respect to some constraints. In many cases, these constraints are unknown to us but we can search and take measurements within an exploration radius. In the machine learning community, we call this safe optimization. One important point in safe optimization algorithms is how fast we can converge to optima and get in a neighborhood of the optimal solution. In this thesis, we introduce a novel safe optimization algorithm that is fast. The experiments are performed through a software package we developed called ASFW which is available for download. On the application side, we demonstrate how safe optimization techniques can be applied to a real-world problem and make an attempt to employ safe optimization approaches to solve a real-world problem and propose a frame work that may be applied in various industrial settings.Applied Science, Faculty ofElectrical and Computer Engineering, Department ofGraduat