This dissertation explores innovative intersections of machine learning (ML) and quantum computing with electron microscopy and quantum simulations, aiming to address and overcome significant challenges in physics and chemistry.First, we introduce a cycle generative adversarial network (CycleGAN) equipped with a reciprocal space discriminator for electron microscopy, which enables the autonomous identification of single atom defects in massive datasets.This method underscores a pivotal advancement towards the automation of materials research by allowing the generation of images that are indistinguishable from real data, thereby facilitating rapid and accurate ML applications in electron microscopy.Furthermore, we delve into the realm of quantum computing to enhance molecular dynamics simulations.By leveraging transfer learning, we train models to predict molecular potential energy surfaces with unprecedented accuracy, utilizing data from Density Functional Theory (DFT) and refining it with output from Variational Quantum Eigensolvers (VQE).This dual-step training I joined The University of Illinois at Urbana-Champaign (UIUC) as an undergraduate in 2012.I will have departed from this institution in 2024.This twelve-year training to become a computational physicist would not have been possible without the incredible support of the people in my life.My journey into the world of physics research began in the summer after my first year of undergrad under the guidance of Professor Karin Dahmen.Her mentorship introduced me to physics research and instilled the confidence to pursue the field further.In the summer after my sophomore year, I transitioned from analyzing slip statistics in nanocrystals to performing simulations and visualizations of binary black hole mergers, marking the beginning of my work with Professor Stuart Shapiro.For the remaining 3 years of my undergraduate and 2 years into my graduate career, Stu pushed me to limits I did not think I could reach, both as a physicist and a computer scientist, and consistently conveyed to me how impactful my work is to his group and the scientific community.In addition to Stu, I had tremendous support and encouragement from my collaborators Milton Ruiz, Antonios Tsokaros, and Vasileios Paschalidis.My experience working with Stu formed my drive to push and demand more of myself.While in the beginning stages of my graduate career, I was particularly interested in the graduate quantum mechanics course taught by Professor Bryan Clark, who first introduced me to quantum computing and quantum information.That class, more than any other, had my undivided attention, as it was the first time I encountered merging the theory of computation with the theory of physics.With many overlapping interests, Bryan welcomed me into his group, where I began to explore the world of developing and applying classical and quantum algorithms to physics.I was free to pursue all my interests with Bryan, and he entertained all my ideas, no matter how bizarre.During this time, I met many collaborators who have helped me shape my ideas into impactful
This dissertation explores innovative intersections of machine learning (ML) and quantum computing with electron microscopy and quantum simulations, aiming to address and overcome significant challenges in physics and chemistry.First, we introduce a cycle generative adversarial network (CycleGAN) equipped with a reciprocal space discriminator for electron microscopy, which enables the autonomous identification of single atom defects in massive datasets.This method underscores a pivotal advancement towards the automation of materials research by allowing the generation of images that are indistinguishable from real data, thereby facilitating rapid and accurate ML applications in electron microscopy.Furthermore, we delve into the realm of quantum computing to enhance molecular dynamics simulations.By leveraging transfer learning, we train models to predict molecular potential energy surfaces with unprecedented accuracy, utilizing data from Density Functional Theory (DFT) and refining it with output from Variational Quantum Eigensolvers (VQE).This dual-step training I joined The University of Illinois at Urbana-Champaign (UIUC) as an undergraduate in 2012.I will have departed from this institution in 2024.This twelve-year training to become a computational physicist would not have been possible without the incredible support of the people in my life.My journey into the world of physics research began in the summer after my first year of undergrad under the guidance of Professor Karin Dahmen.Her mentorship introduced me to physics research and instilled the confidence to pursue the field further.In the summer after my sophomore year, I transitioned from analyzing slip statistics in nanocrystals to performing simulations and visualizations of binary black hole mergers, marking the beginning of my work with Professor Stuart Shapiro.For the remaining 3 years of my undergraduate and 2 years into my graduate career, Stu pushed me to limits I did not think I could reach, both as a physicist and a computer scientist, and consistently conveyed to me how impactful my work is to his group and the scientific community.In addition to Stu, I had tremendous support and encouragement from my collaborators Milton Ruiz, Antonios Tsokaros, and Vasileios Paschalidis.My experience working with Stu formed my drive to push and demand more of myself.While in the beginning stages of my graduate career, I was particularly interested in the graduate quantum mechanics course taught by Professor Bryan Clark, who first introduced me to quantum computing and quantum information.That class, more than any other, had my undivided attention, as it was the first time I encountered merging the theory of computation with the theory of physics.With many overlapping interests, Bryan welcomed me into his group, where I began to explore the world of developing and applying classical and quantum algorithms to physics.I was free to pursue all my interests with Bryan, and he entertained all my ideas, no matter how bizarre.During this time, I met many collaborators who have helped me shape my ideas into impactful
Understanding and simulating strongly correlated electron systems is a long-standing goal for condensed matter physics.The onset of high-performance computers and quantum computing devices provide two powerful playgrounds to simulate quantum systems.In this thesis, I present several new algorithmic methods aimed at simulating quantum systems, both on classical and quantum hardware, in addition to using existing methods to elucidate interesting many-body physics.This thesis would not exist without the guidance and encouragement from my advisor Bryan Clark.His support, even through a global pandemic, was matched with an enthusiasm for physics has made working together a great experience.Bryan's seemingly infinite patience for both confusing topics and career advice were essential in helping me succeed.
This thesis investigates the interplay between machine learning (ML) and quantum mechanics to advance quantum machine learning (QML) and quantum calculations, addressing computational challenges in quantum machine learning, materials science, and condensed matter physics. The high dimensionality of quantum systems often renders traditional methods computationally intensive. By integrating classical and quantum ML techniques, innovative algorithms are developed to enhance efficiency and accuracy. The first part investigates the design and optimization of quantum circuit architectures for quantum machine learning, with a focus on quantum support vector machines (QSVM). I develop a Bayesian algorithm that adaptively constructs quantum kernels, achieving superior performance over classical kernels in small-data classification tasks. I also establish an isomorphism between quantum circuits and polyatomic molecules, which enables the restriction of the search space for compositional optimization of quantum circuits. Additionally, I introduce a flexible scheme that maps quantum circuits into graph representations compatible with the neutral atom quantum hardware. Using this circuit-graph mapping scheme, I develop a Gaussian process regression model with quantum evolution kernels and a classical graph neural network to interpolate QSVM accuracy on graph-structured data. The second part applies ML to accelerate quantum calculations. I develop a variational algorithm based on eigenvector continuation to compute ground state energies of particle-phonon systems. This algorithm constructs the lowest eigenvalue and eigenvector for extended lattices by solving eigenvalue problems for small, independent lattice segments, with a lower computational cost. This enables the application of variational quantum eigensolvers to particle-phonon interactions in large and disordered lattices, reducing the demands on quantum resources. I also developed Gaussian process models to accelerate Green’s function calculations for electron-phonon systems, achieving high accuracy by extrapolating from restricted Hilbert spaces to larger systems with lower computational cost. Ultimately, this thesis demonstrates that these advancements pave the way for applications, including the design of materials, enhanced drug discovery, and the development of scalable quantum computing technologies, thereby positioning the integration of ML and quantum physics as a promising approach for future scientific breakthroughs.
This thesis investigates the interplay between machine learning (ML) and quantum mechanics to advance quantum machine learning (QML) and quantum calculations, addressing computational challenges in quantum machine learning, materials science, and condensed matter physics. The high dimensionality of quantum systems often renders traditional methods computationally intensive. By integrating classical and quantum ML techniques, innovative algorithms are developed to enhance efficiency and accuracy. The first part investigates the design and optimization of quantum circuit architectures for quantum machine learning, with a focus on quantum support vector machines (QSVM). I develop a Bayesian algorithm that adaptively constructs quantum kernels, achieving superior performance over classical kernels in small-data classification tasks. I also establish an isomorphism between quantum circuits and polyatomic molecules, which enables the restriction of the search space for compositional optimization of quantum circuits. Additionally, I introduce a flexible scheme that maps quantum circuits into graph representations compatible with the neutral atom quantum hardware. Using this circuit-graph mapping scheme, I develop a Gaussian process regression model with quantum evolution kernels and a classical graph neural network to interpolate QSVM accuracy on graph-structured data. The second part applies ML to accelerate quantum calculations. I develop a variational algorithm based on eigenvector continuation to compute ground state energies of particle-phonon systems. This algorithm constructs the lowest eigenvalue and eigenvector for extended lattices by solving eigenvalue problems for small, independent lattice segments, with a lower computational cost. This enables the application of variational quantum eigensolvers to particle-phonon interactions in large and disordered lattices, reducing the demands on quantum resources. I also developed Gaussian process models to accelerate Green’s function calculations for electron-phonon systems, achieving high accuracy by extrapolating from restricted Hilbert spaces to larger systems with lower computational cost. Ultimately, this thesis demonstrates that these advancements pave the way for applications, including the design of materials, enhanced drug discovery, and the development of scalable quantum computing technologies, thereby positioning the integration of ML and quantum physics as a promising approach for future scientific breakthroughs.
This dissertation is comprised of works focused on theoretical investigations related to topological crystalline phases of matter.Within this field of study, tremendous progress has been achieved in the past several years in classifying non-interacting fermionic symmetry-protected topological phases, which are characterized by bulk topological invariants.Symmetries of the lattice can be exploited to reduce the complexity of characterizing the bulk nature of topological phases, namely translational symmetry, which is the assumption of a crystal with a perfect, periodic structure, along with internal symmetries (such as time-reversal, particle-hole, and chiral symmetries) and crystalline symmetries (such as rotations, reflections, etc.).The assumption of translational The time I spent at Urbana has served as some of the most formative years of my entire life.I feel incredibly blessed to have had the opportunity to engage in intellectual conversations, start collaborations, and listen to talks from some of the most brilliant physicists at AJL-ICMT.Although there were plenty of rough times during my PhD, some of which were more difficult to overcome than I expected, at the end of the day, I am truly honored to have completed my PhD studies at AJL-ICMT.I will certainly miss Urbana and the countless hours I have spent in the Engineering Sciences Building conversing with colleagues about physics and life, while squeezing in the time to read and write papers, work out calculations, and hastily make presentations.For the truly memorable experiences I have had here, I have many people to thank, all of whom have contributed to my growth not only as a physicist, but as a human being too.First and foremost, I am grateful to my advisor, Taylor Hughes, for allowing me into his research group and taking me on as his student.Taylor showed me how to think independently and creatively and how to recover from repeated setbacks and failures.Most importantly, he encouraged me to have confidence in my own ideas and to pursue them wholeheartedly.Even during the times when I felt that my ideas and the questions I would ask were stupid and worthless, Taylor was able to turn my negative and cynical perspective on my research progress into something positive.He was always incredibly patient with me, willing to explain things over and over again until I finally understood what he was saying.It is rare to find these qualities in an advisor, and although there were times when I felt stranded during this whole process, he always showed me that there was a path forward.If it was not for him, I would not be able to call myself a physicist at the end of this long and hard PhD journey.Second, I would like to thank my collaborators and colleagues through which the years of worth of conversations I have had with them have shaped me into the physicist I am today.I am incredibly grateful to Barry Bradlyn, who effectively co-advised me alongside Taylor during the first three years of my PhD and helped me understand the intricacies of topological quantum chemistry and helped with writing my first paper.I am grateful to Yoonseok Hwang and Giuseppe De Tomasi, two postdocs who joined AJL-ICMT rather late into my PhD but with whom I have had several intellectually stimulating conversations with that have led to plenty of new research ideas and ways of thinking about physics I had not witnessed before.I am also grateful to Penghao Zhu and Oleg Dubinkin, my brilliant group mates for years whose works have inspired me tremendously.They always had the patience and time to entertain my silly ideas and give me plenty of useful feedback on them, and provided me with plenty of help on the simple calculations I needed to get done.
Quantum information science integrates the concepts of quantum mechanics and information science to examine the processing, transmission, and analysis of information.This multidisciplinary field encompasses both theoretical and experimental aspects of quantum physics, aiming to explore the limits of what can be achieved with quantum information technologies.In this thesis, we delve into both the theoretical frameworks and experimental investigations within quantum information science, placing a particular focus on photonic Throughout my Ph.D. journey, I have been fortunate to receive the guidance and support of numerous individuals, each of whom has played an indispensable role in my academic and personal growth.I am profoundly grateful to everyone who has been a part of this remarkable journey.First and foremost, I extend my deepest gratitude to my advisor, Virginia (Gina) Lorenz.Her openness to welcoming me into her research group marked the beginning of an incredibly enriching academic adventure.Gina's mentorship, particularly during my first year in the lab as I worked on my initial experiment, along with her unconditional support and encouragement throughout my Ph.D., has been imperative.Her guidance was especially crucial during the challenging times brought on by Covid-19, providing me with the strength and resilience to continue my research.I am also immensely thankful to my advisor and friend, Eric Chitambar, for his invaluable guidance in the realm of quantum information theory.Eric's teachings, advice, and insights have profoundly influenced my understanding and appreciation of this field.Words fall short of expressing my respect for him, and his impact on my research will undoubtedly extend far beyond my journey
The onslaught of machine learning approaches to many-body problems has demonstrated the power of data-analytic tools in describing phenomena where standard techniques fail.One niche in this new regime of scientific study is topological data analysis (TDA), a field firmly rooted in algebraic topology and only recently explored in the context of quantum information and many-body physics.TDA offers a topological lens to quantum information amenable to direct computation and inputtable to machine learning pipelines.Given its relative lack of application in condensed matter physics, TDA poses a tantalizing route towards observables and insights invisible to traditional approaches.In this thesis we explore applications of this burgeoning field to identify and describe the "shape" of quantum information.Three broad topics comprise our testbed to evaluate TDA's efficacy: many-body localization, relativistic heavy-ion collisions, and multipartite entanglement.While diverse and largely orthogonal, each of these areas of research can be probed by tools from TDA, illustrating its flexibility and versatility.The outline of this thesis is as follows.In Ch. 1, we give an overview of the primary tool of topological data analysis, persistent homology.Through numerous examples we plot out how TDA can be leveraged Grad school is a study in sitzfleisch.By this I mean a lot of sitting, of course, but also a resolve to keep with projects long after the enthusiasm burns away and long before meaningful results appear.It requires a great deal of descending down rabbitholes, ruminating in them, and then quickly exiting before several weeks have passed.Grad school is also a study in via negativa.To learn what questions aren't worth answering, code not worth writing, and projects not worth completing is a failure-induced, ever-evolving skill set.The best lesson I gleaned from grad school is how to recover from failure quickly, as quite late in my career I chose to focus on a niche, relatively unexplored field.With that choice came the terror of finding nothing interesting, coupled with the possibility of profound discovery.There are a lot of people to thank for helping me not fail (which I term a success).Prof. Bryan Clark took me in after I left another research group, and from the beginning he was willing to sit down and read my mangled code and suggest corrections.Prof. Clark could see several steps beyond my half-baked ideas to potential pitfalls, and this foresight certainly smoothed the path.He largely gave me carte blanche to pursue projects that interested me, to which I am grateful.I owe a thanks to my former advisor, Prof. Matthew Gilbert, for accepting me into his research group as a very green first year.My first real research colleagues (Mark Hirsbrunner, Timothy