Colonial Wives: Gender Work and Transcolonial Connections in the British Empire, 1780-1830

DJangi, Parissa State University of New York at Stony Brook ProQue 2019 해외박사(DDOD)

This dissertation uses a particular group of imperial actors to examine the social and cultural strategies of colonialism: the British wives of high-ranking colonial officials. These “colonial wives” accompanied their husbands to sites throughout the empire, where they performed crucial work that enabled the imperatives of the colonial state. I argue that colonial wives’ performances of wifely work helped define racialized notions of British identity during the Age of Revolutions, a critical moment for gender relations, national identity, and imperial transition—changes in colonial rule coincided with added emphases on women’s roles as cultural arbiters and concerns that the manners and behaviors of Britons in the colonies were not sufficiently “British.” This dissertation puts “metropole” and “colony” into the same analytical framework to trace colonial wives’ performances of gender work from Great Britain to Upper Canada, Jamaica, and India. Among their tasks was facilitating imperial patronage, hosting British sociability, and managing colonial households. This work took on new meanings in colonial sites: they deployed wifely work to define a recognizably “British” class of rulers that were culturally, socially, and racially set apart from their subjects.This study adopts a transcolonial perspective to put the Atlantic and Indian worlds into conversation with another and explore the consistent strategies that linked Britain’s colonial projects. By reframing colonialism around everyday acts, my project recovers colonial wives from the margins of imperial history by positioning them and their gender performances as instruments of rule whose focus on regulating the “Britishness” of the ruling class exemplifies the social and cultural reach of the colonial state. I utilize approaches from gender and cultural history as well as performance studies to refine our understanding of the performative, gendered nature of colonialism: feminine acts were gendered technologies of power that the colonial state used across the empire to regulate social and cultural life.

Adversarial Attack and Defense for Image Classification

Wang, Hong State University of New York at Stony Brook ProQue 2022 해외박사(DDOD)

While the deep neural networks have shown impressive performance in many tasks, they are fragile to the carefully designed adversarial attack. To circumvent this vulnerability, numerous defense strategies have been proposed to improve the robustness of DNNs against adversarial examples.This work proposes several adversarial defending methods by focusing on the relationship of the feature maps and representations in the latent space from clean images and adversarial examples. In the first defending algorithm, we use an attention guided knowledge distillation to teach the model to focus on the correct region of adversarial examples, a bi-directional metric learning is then used to constrain the distance between clean image and adversarial examples from two different classes in feature space. The combination of these two modules can improve the robustness of the model by a large amount in both white box and black box settings. Secondly, some research observed that the fragility of deep neural networks may be caused by the non-robust features. In order to extract robust features which are not affected by the adversarial perturbations, i.e., invariant to the clean image and its adversarial examples, we propose a feature disentanglement model to segregate the robust features from non-robust features and domain specific features. We also present some preliminary results for a third defense using contrastive learning.Extensive experiments with various types of attacks on widely used datasets demonstrate the outperformance of our defending methods compared with the state-of-the-art approaches. The detailed performance analysis also shows the effectiveness of constraint and disentanglement of feature space in improving adversarial robustness of deep models.

That Sound in Your Head: Guitar Tone Values in the Entrepreneurial Age

Brounley, David Matthew State University of New York at Stony Brook ProQue 2022 해외박사(DDOD)

This dissertation investigates the ways that guitar sounds are packaged, valued, and folded into musician subjectivity in the early twenty-first century United States. Based on a multi-year ethnographic project at New York city musical instrument store, I argue that the sonic logics that have emerged in guitarist consumer culture are framed by the entrepreneurial values associated with neoliberal ideology. Through interactions in stores, popular media, and in everyday conversations concerning sound, many guitarists consider it a responsibility to produce unique sounds that represent their personality. These composite sounds, usually described as a player’s "Tone," are often likened to sonic fingerprints in both industry literature and casual conversations. Drawing off the importance of this term within guitarist consumer culture, I theorize Tone as a sonic epistemology grounded in the logic and ethics entrepreneurial values.The musical instrument shop is an important fixture in many twenty-first century American musicians’ lives, and yet it has been underexamined by scholars interested in music and capitalism. Indeed, many of the musicians I have interviewed claim to attribute their sonic fluency and self-perceived status as musical connoisseurs through interactions in these spaces, therefore I find it important to give them proper attention. As I recount interactions and interview, I investigate the cosmopolitan discourse-network the sales floor engenders. It is an useful space to study long-standing questions in musicology and music theory, for instance the economization of sound, the linguistics of timbral experience, and community building amongst amateur musicians.

Professional Development for Secondary Science Teachers to Improve Engineering Knowledge, Pedagogy, and Career Advisement

Christian, Kimberly B State University of New York at Stony Brook ProQue 2019 해외박사(DDOD)

This quasi-experimental, mixed-methods observational study reports the design and implementation of an NGSS-aligned engineering workshop series for middle and high school educators focused on the science and engineering practices. The workshop structure and data analysis were based on the interconnected model for professional growth and supported by the situated learning and knowledge integration perspectives on cognition. University-based science, engineering, and science education faculty developed two workshop series in 1) electrical engineering and physics and 2) biotechnology and biology. Each series provided teachers with instruction in the engineering design process, science-engineering integration, and engineering career advisement. In each workshop, teachers engaged in science content-specific engineering activities that were designed to simulate the secondary classroom experience.Science teachers (n=32) voluntarily participated in the workshops and completed pre- and post-workshop surveys. Comparisons of means indicated teachers reported statistically large improvements in their ability to: 1) use engineering activities in the classroom successfully, 2) increase student interest in engineering, 3) help students apply engineering to real-world situations, and 4) advise students on different engineering disciplines and careers, as well as how to prepare students for engineering before college. Pre- and post-workshop interviews with six of these teachers revealed an expansion of teachers’ perceived value of engineering in science and an increased confidence and ability to incorporate engineering and provide guidance to students interested in pursuing a future in engineering. Teachers also expressed their continuing concerns related to the school environment regarding engineering and science alignment.The results from this study demonstrate that a university-based professional development workshop series is an effective intervention to improve the engineering knowledge and skills of secondary science educators, ultimately increasing NGSS adoption in science classrooms. Professional development programs such as this can serve as the foundation for increasing student exposure to engineering and interest in engineering careers.

Study of Particle Correlation and Fluctuation From Nucleus-Nucleus Collisions to Proton-Proton Collisions With the Atlas Detector at the LHC

Zhou, Mingliang State University of New York at Stony Brook ProQue 2019 해외박사(DDOD)

Quark Gluon Plasma (QGP) is a novel state of nuclear matter consist of strongly-interacting quarks. This state only exists in extremely high temperature and density, and can be produced by colliding high energy nuclei. This dissertation utilized several theoretical models and collision data collected by the ATLAS detector at the LHC, to study the particle correlations in both the longitudinal and azimuthal directions.In the longitudinal direction, we proposed a new observable to study the event-by-event multiplicity fluctuation, through decomposing the multiplicity correlation in pseudorapidity into orthogonal bases. The pseudorapidity correlation function contains a significant short-range component, which is estimated and subtracted in a date-driven way. After removal of the short-range component, the multiplicity correlation was found to be dominated by a linear component, and its magnitude was found to be consistent among different collision systems at similar multiplicity. These results provide new constrains on the longitudinal density fluctuation at early stages of the heavy-ion collisions.In the transverse direction, multi-particle azimuthal cumulant technique was applied to study the event-by-event fluctuations of harmonic flow coefficient vn and correlated fluctuations between two harmonics vn and vm. To probe the properties of QGP, we compared the cumulant measurements from two collision systems with different nuclei sizes (Pb+Pb and Xe+Xe), and the results are quantitatively consistent with the hydrodynamical predictions. The influence of centrality fluctuation on the flow cumulants was also studied. This measurement was also carried out for smaller collision systems such as p+Pb and pp. Novel event selection triggers were designed to enhance the event statistics and state-of-the-art subevent algorithms were developed to suppress the non-flow background. These new results provide a handle to disentangle flow fluctuations from the initial and final stages, and also suggest that the prerequisites of QGP might also be achieved in much smaller collision systems.

Towards Studies of the Aubry-Andre Model with Ultracold Atoms in Optical Lattices

Pazmino Velez, Arturo Gregorio State University of New York at Stony Brook ProQue 2019 해외박사(DDOD)

Atomic quantum gases in optical lattices allow for studies of ideal solid state-like systems with precisely controllable system parameters such as interactions, dimensionality and disorder. This thesis is devoted to the development of experimental capabilities, using ultracold rubidium-87 atoms in an optical lattice, to study transport and localization phenomena in one-dimensional lattice systems with weak and vanishing interactions. The first development is the design and implementation of a single-frequency diode laser system at 776 nm suitable for the creation of blue-detuned lattice tubes without residual axial confinement. The general suitability of the laser system for lattice experiments is demonstrated by implementing and studying a superfluid-to-Mott insulator transition. The second development concerns the control of the scattering length for rubidium-87 atoms through a narrow Feshbach resonance at 1007 G. For this purpose, a large portion of the magnetic field system of our apparatus was re-designed, including the coil assembly for magnetic trapping, its high-current switch configuration, and a precise magnetic field stabilization. The capabilities of the new system, in which the current of individual coils can be inverted, are demonstrated by accessing the Feshbach resonance through atom losses, with results that reproduce the state of the art in the field. Moreover, in quadrupole configuration, the new system allows for significant speedup of evaporative cooling.Based on these technical developments, we discuss the feasibility of experiments with incommensurate optical lattices that address properties of a mobility edge in the Aubry-Andre model, as well as of a method for stabilizing Bloch oscillations through a periodic modulation of interactions.

Geometric and Topological Aspects of Self-Assembly: From Spheres and Rods to Designer Particles

Logan, John State University of New York at Stony Brook ProQue 2022 해외박사(DDOD)

In 1959 Richard Feynman gave a lecture at the American Physical Society's annual meeting entitled ``There's Plenty of Room at the Bottom: An Invitation to Enter a New Field of Physics." In the lecture, Feynman imagined the possibility of being able to arrange individual atoms in any geometry we wish. It wouldn't be until the 1980s that self-assembly emerged as the dominant method for realizing this idea. Self-assembly is the autonomous arrangement of molecules into a predetermined structure. Particle geometry, specific interactions, and environmental constraints are just a few of the attributes that act in concert to determine the morphologies available to a set of building blocks. In order to assemble more complex structures, more complex building blocks are required. To better understand how to self-assemble complex structures without error, we must study the different aspects of self-assembly.In this dissertation we highlight fundamental aspects that arise in self-assembly through several examples with increasingly more complex components. We begin with a statistical mechanical description of randomly packed spherical particles, where the average coordination number is treated as a macroscopic thermodynamic variable. The overall packing entropy is shown to have two contributions: geometric, reflecting statistical weights of individual configurations, and topological, which corresponds to the number of topologically distinct states. The theory generalizes concepts of granular and glassy configurational entropies for the case of non-jammed systems.Next, we trade the simplest possible component for the simplest anisotropic particle: rods. We investigate how nanorods interact through a van der Waals force. Rod-like particles are ubiquitous in nature, and are often used in molecular simulations. We seek to find an accurate closed-form, compact expression for the interactions that can be easily implemented in simulations.As we move away from spherical and rod-like particles, the building blocks of self-assembly become more sophisticated, and new methods of analysis are necessary for characterization. In response to this, we present a comprehensive framework of characterizing complex self-assembled structures with a set of orientational order parameters. It is especially relevant in the context of using anisotropic building blocks with various symmetries. A variation of the classical bond order parameters is introduced that takes advantage of the symmetry of constituent particles, and/or the expected crystalline phases. These Symmetrized Bond Order Parameters are demonstrated through applications to (i) a simulated hybrid system that combines isotropic and patchy particles; (ii) experimental data on self-assembly, based on DNA origami.

Cosmic Expansion History Studies from Population Analysis of Gravitational Wave Events

Ye, Jiani State University of New York at Stony Brook ProQue 2022 해외박사(DDOD)

The detections of gravitational wave~(GW) signals from stellar mass binary black hole~(BBH) merger events by Advanced LIGO and Virgo network has inspired the studies of population features of BBH mergers. Being the remnant of star evolution, and widely distributed over space and time, BBH population encodes stellar formation history and cosmic evolution history. In my thesis, I introduce the studies my collaborators and I made on measuring cosmic evolution history with the BBH population features. The abrupt drop in the source frame mass distribution in the Gravitational Wave Transient Catalog~(GWTC) of the first observational run of Advance LIGO and Virgo network indicated an upper limit in the primary mass distribution driven by the Pair Instability Supernovae~(PISN) mechanism. In the first project, my collaborators and I modeled the PISN upper mass limit in the BBH population and studied its constraints on cosmic evolution history with simulated 1-year and 5-year observation data of future observations of Advanced LIGO network with hierarchical modeling. We find that this feature can provide a percent level measurement to the cosmic expansion parameters, comparable to the state-of-art observations of electromagnetic sources. Inspired by the results, I conducted a study on the cosmic expansion measurement based on the simulated observational data of the next generation (3G) GW detection networks. Designed with much higher sensitivities, the BBH catalog of the 3G network are expected to have much larger spatial span and much higher precision, these allow for more features in the mass distribution for cosmic constraints. I constructed two models with different mass features, and found that the cosmic measurement with 1 month of simulated data~($\\sim 10. 4$) are able to constrain cosmic parameters to a high precision, comparable to the future observations of other sources. The measurements are robust over the two models, but potential evolution of the mass distribution features may affect the accuracy of the measurement results. The full population analysis based on hierarchical modeling is computationally expensive, and the constraints on cosmic parameters are mainly from the mass distribution features, thus I explored the possibility of measuring cosmology from the mass distribution feature in more simplified models. These features show correlations with the cosmic parameters, but a failure to include selection effects may lead to a biased inference. The caveats with the high precision measurements of cosmic parameters with future observations in our studies are that, we may have ignored the potential correlations in the event level parameters. Following the studies of metallicity dependence of these parameters, I constructed a phenomenological model including the metallicity correlations, and examined its influence on the cosmic measurements. Depend on the strength of metallicity dependence, our inference can be affected to various extent when failed to model it, but we expect to find substructures in the population feature that help improve our understandings on the underlying correlations.

Infrared and Terahertz Near-field Spectroscopy and Microscopy on 3d and 4d Correlated Electron Materials

Zhang, Jiawei State University of New York at Stony Brook ProQue 2019 해외박사(DDOD)

The electromagnetic waves in the far infrared and terahertz range have ubiquitous applications to the optical characterization of solid-state materials in which numerous physical phenomena occur within the energy range below ~ 100 meV ( molecular rotation, exciton transition, superconducting gap opening, etc.). Conventional infrared and terahertz characterization methods have been suffering from low spatial resolution due to optical diffraction effects. This thesis presents a new type of micro-imaging technique named scattering-type scanning near-field optical microscopy (s-SNOM) which can reach deep subwavelength spatial resolution regardless of the wavelength of probing light. The working principle of far infrared s-SNOM is introduced with an illustration of two experimental measurements on the insulator-to-metal phase transition (IMT) of transition metal oxides Ca2RuO4 and VO2, the images of phase boundaries on sample surfaces show exotic nanoscale phase patterns, revealing the competition between strain and domain wall energy, showing the complex interplay between non-equilibrium electronic and lattice steady states. The second part of the thesis discusses the terahertz time domain spectroscopy (THz-TDS) and optical-pump-terahertz-probe (OPTP) schemes. An experiment is demonstrated to explore the ultrafast electronic dynamics of correlated electron materials V1-xNbxO2 thin films, revealing a novel way of tuning electron-electron and electron-phonon interaction dynamics. The last part of the thesis reports our recent progress in THz s-SNOM with a demonstration of near field imaging on graphene. Measurement suggests that a single layer graphene acts as a perfect terahertz reflector in the near-field regime due to high-momentum effects. Conclusions and potential works for the future are mentioned at last.

Searching for Dark Sectors

Zhong, Yi-Ming State University of New York at Stony Brook ProQue 2016 해외박사(DDOD)

The existence of Dark Matter suggests the presence of a dark sector, consisting of particles neutral under all Standard Model forces. Various "portals" can connect the dark sector to the Standard Model sector. Two popular examples are the vector portal, which gives rise to a "dark photon" ( A'), and the Higgs portal, which gives rise to a "dark Higgs". Such dark forces appear in many well-motivated extensions of the Standard Model. In some cases, they may resolve discrepancies between experimental data and theoretical predictions, such as the muon anomalous magnetic moment. We show that dark sectors and forces can be constrained from several novel probes in current and future experiments, including mono- photon searches in low-energy positron-electron colliders, rare muon decays in the Mu3e, and exotic Higgs decays at the Large Hadron Collider (LHC). We first investigate the power of low-energy, high-luminosity electron--positron colliders to probe dark sectors with a mass below ~ 10 GeV, which couple to Standard Model particles through a low-mass dark mediator. Dark matter candidates in this mass range are well- motivated and can give rise to distinctive mono-photon signals at B-factories and similar experiments. We use data from an existing mono-photon search by BABAR to place new constraints on this class of models, and give projections for the sensitivity of a similar search at a future B-factory such as Belle II. We find that the sensitivity of such searches are more powerful than searches at other collider or fixed-target facilities for dark-sector mediators and particles with masses between a few hundred MeV and 10 GeV. We compare our results to existing and future direct detection experiments and show that low-energy colliders provide an indispensable and complementary avenue to search for light dark matter. We also find that dark photons with masses ~ 10 MeV -- 80 MeV can probed in the rare muon decay process mu+→ e+nue numuA', A '→ e+ e-, in the upcoming Mu3e experiment at the Paul Scherrer Institute without modifying the experimental set-up. We show that the Mu3e has an exciting opportunity to probe a large fraction of unexplored dark photon parameter space, probing kinetic-mixing parameter, epsilon, as low as 10-4 by the end of the experiment. The newly discovered 125 GeV Higgs boson is a scalar and has a small Standard Model decay width. Hence it can easily decay to light dark-sector particles with sizable branching ratios. We investigate several possible dark-sector searches at the LHC based on the exotic decays of the Higgs. We also analyze in detail a possible decay to bbmu +mu- via intermediate scalar states. We find the branching ratio of the exotic Higgs decay can be constrained at the few x10-5 level across a wide range of mass for the intermediate scalars at the high luminosity LHC. Finally, we show a possible dark-sector interpretation of the recently reported 750 GeV diphoton excess from the LHC 13 TeV Run, as the 750 GeV heavy resonance decaying to light dark photons or dark Higgs bosons. Such decays may pass the photon selection criteria and fake diphoton events. We investigate two scenarios where the 750 GeV heavy resonance is a spin-0 or spin-1 particle and estimate the parameters of the models to explain the existing excess.