12023 research outputs found
Sort by
Shocks, Jets, and Emerging Nebulae: Direct Detection and Characterization of Extragalactic Radio Transients in the VLA Sky Survey
Get PDFFor most of their lives, massive stars and supermassive black holes evolve steadily, changing only gradually on human timescales. But on occasion, these cosmic engines erupt, lighting up their surroundings like flashes in the dark. Under the right conditions, the eruptions can manifest as slow radio transients, rising and fading on timescales of weeks to decades. By finding these transients and observing their evolution, we can study otherwise inaccessible aspects of the engines' lives, and piece together their influence on their surroundings.
Until recently, most of our knowledge of slow radio transients came from follow-up observations of explosions first discovered as optical and high-energy transients. This avenue of discovery, while illuminating, provides an indelibly incomplete picture of the radio transient landscape. Moreover, each successful follow-up detection typically comes at the cost of many unsuccessful attempts (measured in both telescope and human time).
My thesis helps address these issues by finding and characterizing transients directly in radio surveys. By applying novel transient detection techniques to data from the Very Large Array Sky Survey (VLASS), I produced the first uniformly selected sample of radio transients associated with local universe galaxies. The 64 transients in my sample have roughly doubled the total number of directly detected slow radio transients in the literature. This sample has enabled the first volume-limited characterization of the demographics of extragalactic slow radio transients. It has also facilitated the discovery of two previously unseen transient types: the merger of a compact object with a massive star and a decades-old emerging pulsar wind nebula.
These early results used only ~15% of the currently available data from VLASS and focused only on extragalactic transients at low redshift. By applying the same techniques to the full survey, I have found ~2000 new transients, increasing the number of known slow radio transients (detected by any means) by a further order of magnitude. With this new sample, my collaborators and I are beginning to shift the study of slow radio transients from the domain of single-object deep-dives to the domain of statistical samples.</p
Multicellular Synthetic Biology in Mammalian Systems
Get PDFIn multicellular organisms, different types of cells use intercellular signals to communicate and regulate population dynamics, and further coordinate complex behaviors. This presents a rarely tapped into potential for mammalian synthetic biology, which was largely restricted to engineering a single cell type in the past to mimic and use similar multicellular designs to achieve more functionalities. However, with current synthetic biology tools and designs, there are several major challenges to achieve a multicellular circuit. Challenges include precise and tunable control over cell type switching, having an orthogonal cell-cell communication signal, and robust control of cell populations.
To address these challenges, this thesis presents a system for tunable regulating of gene expression with DNA methylation, an auxin-based module for mammalian cell-cell communication, and a robust circuit for population control in mammalian cells. I further applied these work to engineering immune cells to show the potential of multicellular circuits in immunotherapies. Together, these works demonstrated the possibility of constructing multicellular circuits in mammalian systems, and that multicellular circuit can further extend the scope of synthetic biology to achieve more complex functions.</p
“It’s Our War Too”: Barriers to Authorship by Women Writing Vietnam War Poetry
Get PDFEven though American women had higher rates of involvement in the Vietnam War than any previous war, poems about their experiences were extremely scarce until over a decade after American troops withdrew. A major contributor to the lack of literary representation is the critical dismissal of women’s war poetry as being unable to teach readers meaningful “truths” about war. This thesis examines two collections of female-authored poems, Visions of War, Dreams of Peace and Shallow Graves, which were published in 1991 and 1986 respectively. The former contains poems from 40 women, most of whom served as army nurses; the latter combines the experiences of Wendy Wilder Larsen, an American woman who lived in Vietnam for two years, and Tran Thi Nga, a Vietnamese woman who immigrated to America. The collections reveal that most American women responded to critical expectations either through self-erasure or active rebellion. In contrast to the American women, Nga is granted authority by critics because her Vietnamese perspective is unique in English literature, but her authorship is instead challenged during the process of adapting her story for an American audience
The "Interpolated Factored Green Function" Method
Get PDFThis thesis presents a novel Interpolated Factored Green Function (IFGF) method for the accelerated evaluation of the integral operators in scattering theory and other areas. Like existing acceleration methods in these fields, the IFGF algorithm evaluates the action of Green function-based integral operators at a cost of O(N log N) operations for an N-point surface mesh. The IFGF strategy capitalizes on slow variations inherent in a certain Green function analytic factor, which is analytic up to and including infinity, and which therefore allows for accelerated evaluation of fields produced by groups of sources on the basis of a recursive application of classical interpolation methods. Unlike other approaches, the IFGF method does not utilize the Fast Fourier Transform (FFT), and it is thus better suited than other methods for efficient parallelization in distributed-memory computer systems. In fact, a (hybrid MPI-OpenMP) parallel implementation of the IFGF algorithm is proposed in this thesis which results in highly efficient data communication, and which exhibits in practice excellent parallel scaling up to large numbers of cores -- without any hard limitations on the number of cores concurrently employed with high efficiency. Moreover, on any given number of cores, the proposed parallel approach preserves the linearithmic (O(N log N)) computing cost inherent in the sequential version of the IFGF algorithm. This thesis additionally introduces a complete acoustic scattering solver that incorporates the IFGF method in conjunction with a suitable singular integration scheme. A variety of numerical results presented in this thesis illustrate the character of the proposed parallel IFGF-accelerated acoustic solver. These results include applications to several highly relevant engineering problems, e.g., problems concerning acoustic scattering by structures such as a submarine and an aircraft-nacelle geometry, thus establishing the suitability of the IFGF method in the context of real-world engineering problems. The theoretical properties of the IFGF method, finally, are demonstrated by means of a variety of numerical experiments which display the method's serial and parallel linearithmic scaling as well as its excellent weak and strong parallel scaling -- for problems of up to 4,096 wavelengths in acoustic size, and scaling tests spanning from 1 compute core to all 1,680 cores available in the High Performance Computing cluster used.</p
The Role of Boundaries and Other Microstructural Features on Emergent Mechanical and Mechanically-Coupled Phenomena at the Nanoscale
Get PDFAs nanotechnology continues to advance, the need for smaller, structurally complex materials has grown. However, these microscopic (10⁶) and nanoscopic (10⁹) structures often display unexpected changes in mechanical properties as compared to their macroscopic counterparts. Nanomechanical studies investigating size-effects in stiffness, strength, recoverability, ductility, and fracture, reveal an intimate interplay between the breakdown in continuum behavior and the energetic landscape of microstructural mechanisms. Additive manufacturing opens new opportunities to explore this microstructure-mechanics relationship as it enables the micro- and nano-scale production of novel materials and microstructures. While existing studies on structural and functional materials highlight the unique size-scale behavior, a large gap remains in our understanding of the complex relationship between microstructure and material performance. This work investigates the interactions and mechanisms that give rise to emergent nanoscale phenomena. With microstructural characterizations, we demonstrate the role of boundaries and interfaces on mechanical and mechanically-coupled behavior in (1) dense nanowire arrays, (2) nano-architected nanocrystalline zinc oxide, and (3) highly-twinned additively manufactured metallic systems. This work provides critical insights into the mechanisms underlying the observed emergent phenomena and further opens our fundamental intuition for microstructure-mechanics relationships in materials at the nanoscale
Computational Investigations of Organometallic Catalysis
Get PDFOrganometallic catalysis facilitates the synthesis of diverse products ranging from polyolefin materials to pharmaceutical compounds, and catalyst performance depends in part on the design of the ligand scaffold. Towards computational ligand design, quantum mechanical methods more fully capture chemical reactivity in comparison to classical methods, but are more computationally demanding. Free energy calculations of key elementary steps of the catalytic cycle permit the computational prediction of catalyst performance and allow modifications of the ligand structure to be explored. In the dissertation, experimental and computational investigations of organometallic catalysis focuses on rational ligand design. Embedding techniques such as embedded mean field theory (EMFT) and quantum mechanics/molecular mechanics (QM/MM) are leveraged in free energy calculations to allow for the reduction of wall-clock times of energy calculations and trajectory sampling. The organometallic systems investigated include Group IV polyolefin catalysts capable of co-polymerization and enantioselective cross-coupling nickel catalysts. Additionally, experimental methodology development is discussed for a nickel-catalyzed cross-coupling of alkynyl nucleophiles to tertiary electrophiles.</p
Dynamics of Protein-Mediated Polymer Coupling and their Implications in Antibody Production and Emergent Patterning
Get PDFProteins serve a wide range of functions in and out of the cell, from signaling and gene regulation to transport and structural reinforcement. These functions are usually carried out from interactions with other molecules in the surrounding medium such as other proteins, small molecules, or DNA. One such class of proteins are what I will call polymer-coupling proteins: these proteins intentionally link identical polymers or two regions of the same polymer together so that their coupled interactions critically affect the state of the biological system. A vast array of such proteins exist in nature with roles such as the looping of DNA to physically inhibit the expression of a gene or the formation of the cytoskeleton which provides a cell with its shape. In this thesis, I use in vitro experimental methods to explore two cases of coupling proteins and understand their roles not only in reorganizing their complementary polymers but influencing the final state of their respective systems.
In Chapter 2, I examine the starting process for the assembly of an antibody-encoding gene in developing immune cells. Motivated by data suggesting that some antibodies are less likely to be made than others, I explore how the early steps of constructing an antibody-encoding gene affect this uneven frequency of assembly. To initiate recombination, the recombination-activating gene (RAG) protein complex simultaneously binds and cuts two well-recognized sequences neighboring two antibody-encoding gene segments in order to allow other proteins to combine these exposed segments together. The sequences to which the RAG protein performs its binding and cutting functions have certain identifiable sequence patterns but can still vary. Through a single-molecule experimental method known as tethered particle motion (TPM) I show how changes to the binding site sequence can enhance or diminish the propensity of the RAG protein to bind and cut the DNA and thus explore the consequences of these altered interactions in the unequal selection for certain antibody gene segments over others.
In Chapter 3, I turn to questions of the emergence of order from self-organization in biological systems. From the molecular to the population scale, biology constantly demonstrates that with an injection of energy, systems can be driven out of equilibrium and allow for the organization of its constituents. A case of such organization in cells is the coupling of microtubules by motor proteins to create and maintain the mitotic spindle, a critical biological architecture for ensuring that each cell obtains a copy of the genome during division. In vitro experiments that exploit similar motor-microtubule interactions have become a convenient way to identify the effects of perturbing a key player such as motor properties or boundary conditions of the system on the spatiotemporal extent of organization. However, in many instances, the dynamics under which such cytoskeletal systems reduce their entropy over the course of creating order have not been carefully examined in experimental systems. Here, I use engineered light-dimerizable motors that can give rise to the formation of a highly connected network that compacts to form a dense, organized structure, and through the use of a noninvasive imaging technique observe how the polymers that make up the network continually reorganize in the bulk during a global contraction of the network.</p
Optimisation & Generalisation in Networks of Neurons
Get PDFThe goal of this thesis is to develop the optimisation and generalisation theoretic foundations of learning in artificial neural networks. The thesis tackles two central questions. Given training data and a network architecture:
Which weight setting will generalise best to unseen data, and why?
What optimiser should be used to recover this weight setting?
On optimisation, an essential feature of neural network training is that the network weights affect the loss function only indirectly through their appearance in the network architecture. This thesis proposes a three-step framework for deriving novel “architecture aware” optimisation algorithms. The first step—termed functional majorisation—is to majorise a series expansion of the loss function in terms of functional perturbations. The second step is to derive architectural perturbation bounds that relate the size of functional perturbations to the size of weight perturbations. The third step is to substitute these architectural perturbation bounds into the functional majorisation of the loss and to obtain an optimisation algorithm via minimisation. This constitutes an application of the majorise-minimise meta-algorithm to neural networks.
On generalisation, a promising recent line of work has applied PAC-Bayes theory to derive non-vacuous generalisation guarantees for neural networks. Since these guarantees control the average risk of ensembles of networks, they do not address which individual network should generalise best. To close this gap, the thesis rekindles an old idea from the kernels literature: the Bayes point machine. A Bayes point machine is a single classifier that approximates the aggregate prediction of an ensemble of classifiers. Since aggregation reduces the variance of ensemble predictions, Bayes point machines tend to generalise better than other ensemble members. The thesis shows that the space of neural networks consistent with a training set concentrates on a Bayes point machine if both the network width and normalised margin are sent to infinity. This motivates the practice of returning a wide network of large normalised margin.
Potential applications of these ideas include novel methods for uncertainty quantification, more efficient numerical representations for neural hardware, and optimisers that transfer hyperparameters across learning problems.</p
Model-Based Lower-Limb Powered Prosthesis Control: Developing and Realizing Nonlinear Subsystem Control Methods for Generalizable Prosthesis Control
Get PDFWhile there are over 600,000 lower-limb amputees in the US, commercially available prostheses remain limited to mostly passive devices. People that walk with a passive prosthesis experience an increase in energy expenditure, a decrease in comfortable walking speed, and gait asymmetry which leads to degenerative conditions. To address these limitations, researchers have developed powered prostheses with the aim of replicating the net positive energy biological limbs supply to humans in walking. These active devices have been shown to decrease users' metabolic cost and increase their comfortable walking speed. However, the control methods to achieve these results typically require hours of heuristic tuning for every user and every behavior. This motivates developing more formal prosthesis control methods that generalize between users.
Formal nonlinear control methods have been developed to realize energy efficient, human-like walking on bipedal robots. These model-based approaches provide a systematic approach to generate and realize provably stable walking gaits. However, these methods cannot be directly applied to prostheses since they depend on a dynamic model of the entire system, and in the case of the prosthesis, the human dynamics are unknown.
To address this challenge, we develop a theoretical framework to translate model-based bipedal control methods to prostheses with the aim of realizing a generalizable prosthesis control method. We separate the prosthesis subsystem from the remaining human portion of the system and model the human's impact on the prosthesis dynamics with a measure of the interaction forces between the human and the prosthesis. We theoretically prove that a model-based controller developed in this separable subsystem framework is equivalent to one developed with knowledge of the full-order human-prosthesis system. With control Lyapunov functions, we develop a wider class of subsystem controllers that solely depend on local information but provide full-order system guarantees, even in the presence of force estimate errors. This work bridges the gap between bipedal control methods and prostheses, allowing us to leverage the benefits of model-based approaches on prostheses.
We demonstrated a controller of this class through an online optimization-based approach on a powered knee-ankle prosthesis, realizing the first model-dependent lower-limb prosthesis controller that accounts for the interaction force between the human and the prosthesis. For a first pass, a force-estimation method was used that yields improved tracking of the desired trajectories over model-independent prosthesis control methods. Then, we incorporated a load cell into the prosthesis platform at the human-prosthesis attachment point to measure the interaction forces, and an inertial measurement to measure the rotation and velocity of the human's thigh. These two sensors completed the prosthesis dynamics model. A pressure sensor incorporated into the prosthesis' shoe measured the ground reaction forces, enabling the prosthesis to respond to its real-world environment, proving robust to 4 different terrains. We extended this controller to a multi-domain hybrid system approach to model the changing contact points occurring in human heel-toe roll. By allowing the prosthesis to sense the human's large varying dynamic load and respond accordingly, this model-based prosthesis controller emulated subject-specific human kinematic trends on a knee-ankle prosthesis for two subjects with no tuning in between, suggesting this approach could yield a method that generalizes between users.</p
Mock Observations of the Sunyaev-Zel’dovich Effect in Massive Galaxy Clusters and a Six-Layer Integral Antireflective Structure for Silicon Optics
Get PDFPart 1: Measuring the kinematic Sunyaev-Zel’dovich (kSZ) effect is a promising observational tool to constrain both cosmic growth and galaxy cluster formation. As millimeter-wave telescopes gain sensitivity and angular resolution over multiple frequency bands, high signal-to-noise imaging of the kSZ effect in large samples of galaxy clusters will become increasingly feasible. However, maximizing the science reach of these upcoming data will require more sophisticated analysis methods to characterize and remove contamination from a range of unwanted signals, such as the emission from dusty star forming galaxies. Current predictions of kSZ-derived constraints do not account for these effects in sufficient detail. Moreover, they typically rely on Fisher matrix analyses, which cannot fully capture the degeneracies among the physical parameters describing the cluster. We present a mock observation and analysis pipeline to determine the science reach of kSZ galaxy cluster observations that employs more detailed noise models and more sophisticated analysis methods. From our mock observations, we derive new forecasts of the constraining power of next-generation telescopes on cluster peculiar velocities for several instrument configurations from the 10-m, 30-m, and 50-m classes. These forecasts will inform the designs of next-generation telescopes targeting kSZ observations and will indicate the optimal instrumentation for both cosmological and cluster-scale constraints. The software pipeline we develop will also be directly usable as an analysis tool once observations from such telescopes become available.
Part 2: Silicon optics can greatly benefit future millimeter and submillimeter astronomical instruments thanks to silicon’s useful properties such as low loss, high refractive index, and high strength. However, silicon’s high index (n = 3.4) necessitates antireflection (AR) treatment, which has proven a major challenge, especially for the multilayer treatments required for wide spectral bandwidths. We present our approach to this challenge, in which we develop a wide-bandwidth integral AR structure for silicon optics that uses a novel fabrication technique that combines deep reactive ion etching (DRIE) and wafer bonding. We have previously demonstrated a two-layer AR structure for windows over a 1.6:1 bandwidth and are currently fabricating a four-layer coating for a 4:1 bandwidth. Here, we focus on a design for a six-layer structure optimized to give -20 dB reflection between 80 and 420 GHz (5.25:1 bandwidth), which will be useful for future multicolor SZ observations.</p