Nonlinear optics has long been studied as a basis for realizing fast information processing devices and sensors. Nonlinear photonic logic enables the elimination of slow electron-photon transduction processes and supports high signal bandwidth at optical carrier frequencies. Recent breakthroughs in material science have ushered in a new era of research on atomically thin 2D materials with strong and novel properties for nonlinear optics. 2D materials may be incorporated with scalable nanophotonic technologies via post-lithographic integration. This thesis presents fundamentals for nonlinear optical properties of the 2D materials and their application for nonlinear quantum optics. It first describes the electronic band structures of two prominent 2D materials, namely, graphene and monolayer MoS2. Their linear and nonlinear optical properties are analyzed and presented. The analysis of graphene's optical property uses an innovative first-order perturbative S-matrix formalism that systematically identifies various physical mechanisms contributing towards the optical Kerr nonlinearity. On the other hand, the analysis of the optical property of monolayer MoS2 adopts a massive Dirac Hamiltonian that leads to linear and nonlinear optical susceptibilities through a standard perturbative calculation. It turns out that, although its real Kerr nonlinear susceptibility is enormously large compared to bulk materials, graphene has an even larger imaginary Kerr nonlinear susceptibility that degrades coherence via strong two-photon absorption. Hence, graphene is not a suitable material for applications where coherence is essential. In contrast, monolayer MoS2 that has non-zero finite bandgap energy turns out to be a simultaneously suitably coherent and highly nonlinear material as its real-to-imaginary ratio of the Kerr nonlinear susceptibility can be adjusted through detuning the optical carrier frequency. The thesis also presents a metamaterial configuration based on Kerr nonlinearity of the monolayer MoS2 coupled to a local surface plasmon. The unique combination of a strong field enhancement from the plasmonic effect and the atomic thickness of highly nonlinear 2D materials constitutes an optical nonlinear oscillator. This system produces highly quantum behavior, namely, photon antibunching and non-Gaussianity. When built on rapidly developing nanophotonic platforms, 2D materials are promising nonlinear optical materials that have a vast potential for a future large-scale quantum information processing network.

This dissertation examines the visual culture of the Wuyi Mountains, one of the sacred Daoist sites in China's southeastern Fujian province. Since the twelfth century, the breathtaking topography of Wuyi has inspired a great variety of cultural productions, including stone inscriptions, topographic paintings, cartography, and contemporary theatrical performances. Grounded in the cultural practice of boating initiated by the influential Confucian philosopher Zhu Xi (1130-1200), these cultural productions developed amid Confucian, Daoist, tourist, and contemporary political currents. Reflecting diverse ways in which humans engaged with Wuyi, these visual materials have defined and augmented the mountains' sacredness. Through integrated analyses of ecology, topography, and visual culture, this study traces the dynamic yet neglected ecological episodes of Wuyi. I argue that understanding ecologically engaged artistic production illuminates historical conceptions of the sacred, reveals histories of human intervention in and interaction with Wuyi, and exemplifies the ways in which we can re-imagine our relations with the environment.

Maximum likelihood estimation (MLE) is influential because it can be easily applied to generate optimal, statistically efficient procedures for broad classes of estimation problems. Nonetheless, the theory does not apply to modern settings—such as problems with computational, communication, or privacy considerations—where our estimators have resource constraints. The thesis will introduce a modern maximum likelihood theory (through generalization of local minimax theory) that addresses these issues, focusing specifically on procedures that must be computationally efficient or privacy-preserving. To do so, I first derive analogues of Fisher information for these applications, which allows a precise characterization of tradeoffs between statistical efficiency, privacy, and computation. To complete the development, I also describe a recipe that generates optimal statistical procedures (analogues of the MLE) in the new settings, showing how to achieve the new Fisher information lower bounds.

Through recent works, I have been interested in the perceptual transitions between intelligibility and ambiguity with language in music. With Adhocracies I wanted to attempt a much larger scale transition between the states of unintelligible chatter, to ambiguous speech, to (finally) articulated prose. Adhocracies has become a vessel where I could address these language and speech procedures in the same context as several other compositional devices I have been developing concurrently.

My research with the microfluidic Reservoir-on-a-Chip (ROC) platform has produced multiple engineering science contributions toward investigating the fundamental mechanisms that dictate transport through subsurface porous media. Microfluidic devices, better known as micromodels, are devices with a connected porous network that allows the direct visualization of complex fluid flow dynamics occurring under transient conditions. The porous pattern of micromodel in my study is analogous to that of natural reservoir rock (i.e. sandstone or carbonate). The micro-pattern is etched in a crystalline silicon wafer with the DRIE (deep reactive ion etching) technique which offers a large aspect ratio (i.e. pore throat-to-body ratio), with more realistic and well-defined structures. Consequently, investigating fluid flow through representative pore network patterns and material in the micromodels have been greatly beneficial to petroleum, geologic, and environmental engineering field. Micromodel studies are based on the direct observation of the pore-scale fluid structures, the visualization of the flow field, and the characterization of matrix-fluid and fluid-fluid interactions. I implemented various methodologies that enable the real-time monitoring of events occurring in a micromodel by integrating them with high-resolution microscopy and laser-induced fluorescence. My research improves petrochemical and geophysical characteristics of transports in micromodels through the development of new micro-fabrication processes, new experimental frameworks, imaging, and novel image processing algorithms. First, my research addresses greater realism in pore structure and visualization of micromodels for the characterization of single and multiphase flows. I optimized dual-etching fabrication and improved 3D structural realism of carbonate-like flow networks inside the micromodel. I applied the micro-particle image velocimetry (micro-PIV). The micro-PIV provides insights into the fluid dynamics within microfluidic channels and relevant fluid velocities controlled predominantly by changes in pore width and depth. Compared with conventional single-depth micromodels, micro-PIV and fluid desaturation pattern prove that the dual-depth carbonate micromodel is a better representation of pore geometry showing more realistic fluid flow and capillary entry pressures. Second, I demonstrated, for the first time, that micromodels monitored using advanced spectral imaging enables real-time and in-situ quantification of the local viscosity of non-Newtonian viscoelastic polyacrylamide EOR polymers. This, in turn, paves the way to validate computational fluid dynamics models for viscoelastic fluids. Third, novel deep-learning algorithms (convolutional neural networks) were applied to the micromodel images for the automated analysis of surface properties. With proper training of deep-learning architectures on high-quality image datasets, I proved that deep-learning has a great potential to serve as a quick and automated image analysis tool for surface wettability determination with an accuracy larger than 95%. Forth, I established an in-house micro-fabrication procedure using a Direct-Write-Lithography technique for the rapid prototyping of new microfluidic designs. I worked on optimizing the micromodel channel design to make the micromodel more suitable for direct visualization of micro-pore scale mixing dynamics between precipitant and oil phase, which may cause asphaltene aggregation and their agglomerations. Furthermore, confocal microscopy enables the 3D reconstruction of asphaltene agglomerates; it reveals the size and size distribution of asphaltene aggregates as a function of flocculation time.

Cells are the smallest functional unit of the human body, which contains 37 trillion cells on average. Current technology for intracellular sensing does not yet allow us to follow biological processes inside living intact cells on a continuous, real-time basis. Information about such processes is invaluable for understanding complex diseases, such as cancer, immune-pathological, and cardiovascular disorders. For example, the acidity in the micro-environment inside cancer cells can undergo transitory stages that last up to a few hours and are difficult to detect by non-continuous detection methods. However, most of today's diagnostic tests and medical monitoring operate at the body, organ, or tissue level. Continuous detection of cellular activities can significantly advance our knowledge of biology and disease and enable new discoveries in early diagnostics and tailored therapeutics. This dissertation presents the exploratory work on a new approach toward intracellular sensing with chip-in-cell (CHIC) devices. Each CHIC device has a wireless front-end to communicate data and a sensor interface responsive to a specific stimulus. We present the design, fabrication, and characterization of the wireless front-end with a radio-frequency identification (RFID) and a corresponding external detector. The RFID diameter is 22 μm and can be naturally internalized by certain classes of living cells. This is the world's first cellular-implantable RFID. The detector can identify RFID presence and distinguish RFID capacitance changes via resonance frequency shifts. Requirements on system-level integration with biological microfluidic platforms are included. To detect biologically meaningful parameters, a pH-responsive hydrogel is integrated as an example prototype CHIC sensor interface. The hydrogel is fabricated onto interdigitated electrodes. Capacitance shifts at gigahertz range are characterized in physiologically relevant pH ranges from 6 to 9, with steps as small as 0.25 pH identified. Overall, the CHIC sensors are addressable to individual cells, bio-compatible for long-term observation, and sensitive to specific chemical changes; a solid step toward continuous real-time intra-cellular sensing.

In this thesis we discuss designing fast algorithms for three problems in computational genomics. 1) Genome Assembly: With the proliferation of DNA sequencing and decreasing prices, genome assembly is the fundamental problem in computational genomics. Every organism has a genome, which can be abstracted as a long string of A, C, G, T. The length of the genome varies from millions in bacteria to hundreds of billions in some plants. We currently do not have the technology to read genomes end to end and just get short-noisy substrings from the genome called reads. The lengths of these reads are around 10k and error rate is around 15%. Reconstructing the genome from the reads is the problem of genome assembly. The conditions for being able to recover the genome exactly depends on the repeat structure of the genome and are explored extensively in literature. But in practice, we find that there are many datasets where this is violated. We develop algorithms that can optimally recover the genome to within informationally possible limits given a set of reads. We also implement this in the form of an open source software: HINGE, which is used widely in bacterial genomics. This work is joint work with Ilan Shomorony, Fei Xia, Tom Courtade and David Tse. 2) Haplotype Assembly: This is a problem associated with reference based Sequence assembly. Humans are diploid: that is, they have two copies of each chromosome. Roughly speaking, the two chromosome are identical on all apart from a few positions (around 1\% of positions) called single nucleotide polymorphisms or SNPs. In this case reads can be substrings of either chromosome which is not known, and one wants to infer the sequence of SNPs on each chromosome. We show that this problem can be seen as decoding a convolutional code, and reduce it to a graph clustering problem. We then come up with a spectral machine learning algorithm to solve the problem. We show derive tight estimates of the number of measurements necessary theoretically. This work is joint work with Eren Sasoglu, Yuxin Chen and David Tse. 3) Speeding up Machine learning algorithms: The celebrated Monte Carlo method estimates a quantity that is expensive to compute by random sampling. We propose adaptive Monte Carlo optimization: a general framework for discrete optimization of an expensive-to-compute function by adaptive random sampling. Applications of this framework have already appeared in machine learning but are tied to their specific contexts and developed in isolation. We take a unified view and show that the framework has broad applicability by applying it on several common machine learning problems: k-nearest neighbors, k-medoids, hierarchical clustering and maximum mutual information feature selection. On real data we show that this framework allows us to develop algorithms that confer a gain of a magnitude or two over exact computation. We also characterize the performance gain theoretically under regularity assumptions on the data that we verify in real world data. We stumbled onto this problem and approach while trying to run k-medoids on a single-cell RNA-seq dataset. This is joint work with Vivek Bagaria, Tavor Baharav, Vasilis Ntranos, Martin Zhang, and David Tse.

My dissertation is composed of three articles, focusing on student-centric changes in American higher education in the post-World War II era. The first article illustrates the expanding notion of students as "whole persons" in the scholarship of higher education and the concomitant rise of student affairs as a professional field. The second article examines the current state of student centrism in American higher education reflected in mission statements. The third article explores the relationship between organizational characteristics and a university's discursive, structural commitment to diversity, inclusion, and equity. More specifically, Article 1 explores the ways in which student centrism and student empowerment in American higher education has arisen in the past few decades. With a descriptive analysis of the historical development of keywords in journal articles, the author first demonstrates how sociological and educational scholarship reflects the shifting cultural view on students, increasingly understood as persons, and student life, increasingly complex and multidimensional. This article also illustrates how the field of student affairs has developed, in conjunction with the rise of various dimensions of student life in academic research. The impressive expansion of student-focused research and the structuration of the field of student affairs is discussed with a macro-sociological perspective. Drawing on the neo-institutional approach, this article seeks to explain such rise with two cultural principles of modern society: (1) the valorization of science and rationalization and (2) the expanded notion of personhood. The findings suggest that, beyond market competition, there is a widely held belief that student life is to be managed comprehensively and systematically, by professionals with formal training. While Article 1 focuses on the scholarship of higher education, Article 2 examines the behaviors of higher educational institutions. Specifically, this article explores the university's presentation of various themes of student life in mission statements, using a nationally representative sample of 218 non-for-profit public and private institutions. With increased competition and uncertainty, American higher educational institutions face greater pressures to manage their public image and identity. The findings of a descriptive analysis reveal that virtually all universities and colleges have adopted a mission statement and that there is no systematic pattern across various types of institutions. American higher educational institutions tend to emphasize students' holistic experience, the functional purpose of higher education, learning as well as research. The predominant themes and patterns in mission statements of the random sample are compared with those of the most selective institutions in the United States. Lastly, with historical examples, the author illustrates how and why the form and content of the mission statement has been homogenized over time. Article 3 pays special attention to the issue of diversity. With intensifying discussion on student empowerment and student diversity in higher education, a growing body of literature examines educational benefits of campus diversity for various outcomes ranging from student development to institutional excellence. This article focuses on organizational aspects of diversity initiatives in contemporary American higher education. Specifically, the study explores the relationship between various institutional characteristics and the university's discursive as well as formal structural commitment to diversity, using logistic and zero-inflated negative binomial regressions. Based on a uniquely compiled dataset of organizational characteristics and mission statements for a cross-sectional sample of 236 four-year non-for-profit colleges and universities across the United States, the findings suggest that cultural and institutional factors, above and beyond functional needs, influence (1) a university's emphasis on diversity in its mission statement, (2) its establishment of the diversity office, and (3) its appointment of the high-rank diversity officer. Differential relationships are found for discursive and formal structures, indicating loose coupling in the higher education system. The implications are discussed.

Performance-based earthquake engineering (PBEE) has in many ways revolutionized the thinking about seismic engineering design and acceptable performance of buildings in earthquakes. It is now making its way into commercial engineering design and risk analysis practice, as engineers aim to design better-performing buildings, and holders of mortgage or insurance instruments try to better understand the risk they face from damage to associated buildings. Some parts of the calculations (e.g. structural response) have been extensively assessed and validated. There are few similar studies, however, that focus on the damage and loss predictions. The purpose of this dissertation is to address this, by analyzing, evaluating, and improving the damage and loss predictions. The specific PBEE methodology examined in this dissertation is the FEMA P-58 Seismic Performance Assessment Procedure. FEMA P-58 damage and loss predictions are analyzed, to determine how they are impacted by other parts of the calculations. Firstly, variance-based sensitivity analyses are conducted to investigate the interaction of loss predictions with different inputs to the calculations. Of the six inputs considered in the analyses, it is found that predictions of building repair cost (as a fraction of replacement value) are most sensitive to shaking intensity and building age, while building re-occupancy time predictions are most sensitive to shaking intensity and building lateral system. Secondly, a methodology is developed to quantify the impact of available structural response data from seismic instrumentation on the quality of the damage and loss predictions. The density of instrumentation examined using the methodology ranges from the case in which all floors are instrumented to that in which no floors are instrumented and simplified procedures are used to produce structural response predictions. It is found that the quality of the predictions generally improves as the density of seismic instrumentation increases, but it is not crucial for the density to be very high to achieve reasonable accuracy in both damage and loss predictions (although this may depend on the arrangement of instrumentation within a building). Loss predictions are evaluated using data observed in previous seismic events, to understand the degree to which they reflect real-life consequences of earthquakes. A methodology is developed for evaluating the ability of FEMA P-58 component-level losses to predict damage observed for groups of buildings. It is found in applications of the methodology that FEMA P-58 non-structural component-level loss predictions provide more insight into damage than variations in ground shaking between buildings. Finally, this dissertation includes a number of recommendations for improving non-structural mechanical component fragility functions and associated loss predictions used in FEMA-58 calculations. The fitting technique currently used for the functions does not converge in some cases, and the methodology used to predict anchored mechanical component losses can lead to some unexpected results, such as non-smooth variation of repair costs with anchorage capacity. An alternative statistical technique is proposed for fitting the fragility functions that mitigates the non-convergence problems when fitting and makes predictions that better align with damage observed in past events. A more intuitive methodology for predicting anchored mechanical component losses is also suggested. The findings of this dissertation help to enhance understanding of, and improve, the damage and loss predictions used in the FEMA P-58 seismic performance assessment procedure. They ultimately enable various stakeholders, such as building owners, design professionals, lenders, and insurers, to make more informed decisions about seismic risk.

Utility ratepayer-funded energy efficiency programs are the prevalent method for reducing the energy consumption of single-family homes. These programs rely on estimating the electric energy efficiency savings potential of various EEMs (e.g., added insulation or weatherization) to determine their cost-effectiveness, which is the main criterion for determining the success of a program. Program managers lack a cost efficient and evidence-driven method to infer the energy savings potential of a specific energy efficiency measure in a single-family home. This dissertation presents a cost-efficient and empirical evidence-driven method for inferring the energy savings potential of an energy efficiency measure by analyzing moments when the inefficiency of the relevant characteristics is most prominent. This Inefficiency Estimation Method selects homes using only readily available information, specifically 15-minute electric utility data (i.e., interval data) and hourly ambient weather data. The cornerstone of this method is the application of an objective Bayesian model. The use of Bayesian statistics has several benefits over classical statistical methods. One benefit is the ability to improve the accuracy of the analysis by adding new information into the model through informed objective priors. I compared the results of my Inefficiency Estimation Method to selecting homes based on annual energy consumption and selecting homes based on annual A/C energy usage, the currently available empirical methods. The comparative metrics for the methods were energy savings and cost-effectiveness. The inefficiency estimation method calculated 2X more energy savings and 30% better cost-effectiveness for A/C unit replacement than the existing methods. For selecting homes that need insulation, the inefficiency estimation method calculated 18% more energy savings and approximately the same cost-effectiveness as compared to the A/C energy usage method. These results show that it is feasible to target homes by inferring the energy savings potential of specific energy efficiency measures with interval data and weather information using Bayesian inference. This selection method provides energy efficiency program managers an empirical evidence-based way for household level targeting that has the potential to increase such programs' energy savings and cost-effectiveness. Future work can investigate the ability of the Inefficiency Estimation Method to infer the energy savings potential of other energy efficiency measures.