Dissertation Committee Chair: Prof. Zohreh Davoudi
Committee:
Prof. Ian Spielman
Prof. Jacob Taylor
Prof. Norbert Linke
Prof. Xiaodi Wu
Prof. Justyna Zwolak
Abstract: Theoretical physics research often requires the most advanced mathematical techniques, ranging from geometry and calculus to numerical methods and data-driven techniques used on high-performance computers. As one of the most complex and demanding sub-domains of physics nowadays, quantum information science challenges scientists to leverage and expand our toolbox in order to analyze and comprehend quantum mechanical systems. This dissertation provides a quick overview of the various hardware platforms and describes an analytical toolbox implemented in quantum information science. This toolbox contains the three most used methodologies in modern scientific research: analytical modeling, numerical simulation, and machine learning. We address historical contexts, technical specifics, application examples, and the advantages and downsides of each approach. We focus on three case studies to show the usage of this integrated analytical toolbox: First, as a qubit noise reduction study, we describe a giant atom dissipation process using a superconducting qubit coupled nonlocally to a mechanical waveguide; Second, as a quantum control optimization problem, we design a feedback cooling strategy based on a toy model of weakly monitored Bose-Einstein condensates. At last, as an image data analysis, we employ neural network architectures to classify and analyze our experimental dataset automatically.