71446 research outputs found
Sort by
Dynamic, Welfare-Maximizing Pooled Testing
No full textFrom isolated cases to global pandemics, testing for disease is fundamental to modern medicine. However, shortages of time, testing supplies, and staff means that it is often not possible for each biological sample to be individually tested. Thus, pooled testing is often used by the American Red Cross, medical clinics across the United States, and health centers throughout the Global South. Under pooled testing, biological samples are grouped together and tested as one sample. In this thesis, I first build upon the work of Finster et al., who developed a general model for this problem. I first provide practical implementations for the algorithms theoretically presented there, and develop methods to determine the posterior probability of health given a history of test results. I then introduce novel algorithms to improve pooled testing assignments by using a dynamic approach. In this setting, tests are sequentially assigned, increasing the expected welfare of pooled testing by conditioning on the results of past tests before determining the next. I develop and discuss four algorithms under the dynamic setting: one that computes the optimal dynamic testing allocation, a quicker but less optimal algorithm that uses a greedy heuristic, and two classes of machine learning-based algorithms, one using supervised learning and one using reinforcement learning. I evaluate these algorithms by their computational complexity, optimality guarantees, and performance across various testing instances. My results show that the benefits of dynamic testing are significant, especially under real-world constraints. While the optimal dynamic testing allocation is welfare-maximizing, the computational complexity of the algorithm makes it infeasible for practical use. The greedy dynamic-assignment algorithm yields an increase in welfare of over 2.75% compared to the previous state-of-the-art algorithm for large populations and outperforms this benchmark in over 53% of instances. While the performances of the machine learning-based algorithms are not currently competitive, the results suggest promise and motivate further work on such an approach. I also use the empirical results to discuss how various changes in constraints, such as testing budget, affect the variance of welfare yielded under dynamic plans. The results show that current pooled testing practices can be meaningfully improved through dynamic assignment, particularly through the greedy dynamic-assignment algorithm. Adopting the proposed methods would result in a variety of welfare-increasing outcomes, from family members being able to visit immunocompromised loved ones more often to an increased number of life-saving blood donations delivered to patients in need.Applied Mathematic
Warping the Weft: Ecological Interstices Through Material Compositions
No full text“Warping the Weft” examines the potential of public art to create unique ecologies as a distinctive public artwork. Often positioned as two separate disciplines, this project explores the emerging possibilities of integrating landscape architecture and public art. Utilizing the Percent for Art program, a public artwork is integrated into the redevelopment plan of the Kingsboro Psychiatric Hospital in East Flatbush, New York. Using the format of a weave, mosaic configurations of material and plants create portals revealing historical realities of the site: the oak-hickory forests of Lenapehoking, colonial farmlands, and remnants of the former psychiatric hospital built for the rapidly urbanizing city. Glass imagery is integrated into the mosaic, as speculation on possible past and future landscapes within the site. Water acts as a connective material: swales empty and fill to complete the weave, speaking to ever-present yet unaddressed perspectives and emotional realities omitted from historical records.Department of Landscape Architectur
Aspects of Symmetry for Flat Space Holography
No full textA central goal of modern high energy theory is to understand the structure of holographic dualities
beyond asymptotically anti-de Sitter spacetimes. In asymptotically flat spacetimes, holographic dualities
are expected to relate scattering amplitudes to correlation functions in a lower-dimensional theory living
on the celestial sphere. However, it has proven challenging to construct examples of boundary theories
with the full symmetry group of quantum gravity in flat spacetimes. In this thesis, we study how these
bulk symmetries can emerge from lower-dimensional conformal field theories. After extending the celestial
dictionary to (2,2)-signature Klein space, we show that fully Poincaré-invariant scattering amplitudes can
emerge from building blocks of 2D CFTs. Next, we study how breaking translation invariance modifies
the dual theory. We relate gluon scattering amplitudes in the presence of a translation-invariance breaking
source to fundamental building blocks of the AdS/CFT correspondence. We then show that these translation-
breaking backgrounds centrally extend the soft algebra of the dual theory, and use this to generate toy
examples of celestial dualities for self-dual Yang-Mills with broken translation invariance. Finally, we
study novel ways for Poincaré invariance to emerge from a lower-dimensional CFT. After showing that
translation-invariant theories can emerge from entangled states in theories without translation invariance,
we demonstrate that translation-invariant scattering amplitudes can be extracted from linear combinations of
smooth CFT correlation functions. Ultimately, we construct an example where gluon scattering amplitudes
can be extracted from a simple lower-dimensional theoryPhysic
VR-chaeology: Assessing the Potential of Virtual Reality in Demonstrating Archaeological Processes for Enhanced Learning in Museum Exhibits
No full textAdvances in alternate reality technologies (virtual reality/augmented reality) have caught the attention of educators in museums and cultural heritage sites as potentially useful interactive learning tools. However, there is a lack of data-driven studies on the usefulness of these technologies or their overall feasibility. This thesis provides a case study that quantifies the learning and enjoyment of museum visitors when a VR component is added to a more traditional museum exhibit. The virtual reality integration gives users an opportunity to engage in flintknapping, or the process of making stone tools. This classic form of experimental archaeology was selected because, while it is a complex process that was commonly employed in the past, it is not typically utilized today, and therefore provides what is likely a novel experience for visitors. Flintknapping is also a critical index of past human cognition. This thesis tests and compares the learning outcomes of the exhibit both by itself and with AV and VR components. Results show that increasing levels of interactivity correlate with both better learning outcomes and increased enjoyment but does not correlate with desire to learn more about the topic. Finally, this thesis assesses and discusses the feasibility of VR for the traditional archaeological museum.Computer Scienc
Lost "Reality": Beyond Conservation Against Deforestation
No full textLost Reality examines the accelerating deforestation in the southern Amazon, driven by extensive road networks, extractive logging, and ineffective conservation policies. Centered on the construction of the BR-319 highway, a destructive corridor cutting through the forest, the city of ‘Reality’ has emerged as a focal point for new deforestation activities.
In response to the failure of conservation policies, the work draws upon communal ways of living within the forest reimagining conservation through strategic design interventions. It proposes a socio-environmental approach that recognizes the Amazon as an inhabited landscape of humans and non-humans. This new occupation framework transforms road axes into a conservation network that reframes logging and ecological stewardship as a living system. Lost Reality envisions systems integrating diverse inhabitation practices, evolving with the forest’s temporal cycles to imagine a more resistant reality for the Amazon forest.Department of Landscape Architectur
Existential Distress in Palliative Care: Age-Related Differences and Provider Perspectives
No full textThis study centers around existential distress in the palliative care setting, a common phenomenon observed during life-threatening illnesses, such as cancer. The aim of the study was to explore how existential concerns and distress manifests across advanced cancer patients, how it is identified and assessed for severity, and how developmental differences (between older and younger patients) influence its presentation.
Eleven palliative care providers from cancer centers across the United States participated in this study. The sample included both inpatient and outpatient practitioners from sites with robust palliative care programs. This qualitative study utilized semi-structured interviews to gather insights on the nature of existential concerns and distress in their patient population.
The findings revealed universal manifestations of existential and patient concerns, irrespective of a patient's age. However, there were differences in the way those manifestation was expressed across age groups. Key time points for identifying existential distress were identified (acute prognosis, when role shifts and losses occur, and when behavioral changes occur), and developmental differences in existential domains and patient concerns were explored. The unexpected finding of countertransference and self-identification of providers with younger patients was also examined.Extension Studie
Judging Fairness, Fueling Conflict: The Role of Personality and Bias in Online and Offline Aggression
No full textNumerous divisive profiles on social media contribute to disinformation, hate speech, and the recruitment of individuals into extremist beliefs. Unlike past eras, today’s influence is global, resulting in minimal restrictions on viewing and amplifying harmful content.
This study uniquely examined people’s connections— and disconnections—with others by focusing on beliefs about deservingness, the (mis)use of social media, and trolling behavior. No prior research has integrated these three elements to enhance people’s understanding of deservingness beliefs in social and online contexts.
The primary aim of this study was to develop a measure of divisive identity and create a potential profile of individuals exhibiting divisive traits. This measure could inform strategies to address the challenges posed by such behaviors in the modern interconnected world.
Keywords: deservingness, divisive traits, social media, trolling behaviorExtension Studie
Theorizing Pathways to Family Formation: Birth, Adoption, and Infant Abandonment Under China’s One-Child Policy
No full textChina’s one-child policy, which restricted the number of children allowed per family from its inception in 1979 until its conclusion in 2015, has attracted much scholarly attention. Sociologists have documented the policy’s various impacts, from the aging workforce and skewed sex ratio to the effects on women’s reproductive health and state-society relations. This dissertation examines one understudied impact of the one-child policy: the abandonment of female infants. Caught between state demands for a smaller family and familial demands for a son, and with no legal means to place children for adoption, thousands of Chinese families left their female infants anonymously in public places. Some children died before being found, while many were reabsorbed into Chinese society via informal, unregistered adoptions. The remaining unclaimed children were funneled into state orphanages, from which at least 150,000 were sent abroad to families in North America and Europe.
Taking advantage of the author’s unprecedented access to families who abandoned daughters, this dissertation analyzes the context, process, and aftermath of infant abandonment from the perspective of the families themselves. The dissertation is composed of three essays that draw from ethnographic fieldwork in rural China and in-depth interviews with families who abandoned children and comparison families who did not. The first essay asks how states and institutions structure cultural aspirations for an ideal family; the second essay investigates the cyclical process of striving to realize this ideal family; and the third essay examines the emotional aftermath of infant abandonment as one particular cost of striving. Overall, this dissertation presents family formation as a socially patterned process, consequential for our understanding of aspirations, inequality, and grief.Sociolog
An opponent striatal circuit for distributional reinforcement learning
No full textMachine learning research has achieved large performance gains on a wide range of tasks by expanding the learning target from mean rewards to entire probability distributions of rewards — an approach known as distributional reinforcement learning (DRL). The mesolimbic dopamine system is thought to underlie reinforcement learning in the mammalian brain by updating a representation of mean value in the striatum, but little is known about whether, where, and how neurons in this circuit encode information about higher-order moments of reward distributions.
In this dissertation, we first review the mathematical foundations and biological plausibility of these DRL algorithms. We then provide the first experimental and computational evidence that the striatum, specifically, learns such reward distributions. We used high-density probes (Neuropixels) to acutely record striatal activity from well-trained, water-restricted mice performing a classical conditioning task in which reward mean, reward variance, and stimulus identity were independently manipulated. In contrast to traditional RL accounts, distributions with the same mean and variance were encoded more similarly than those with the same mean but different variance. Remarkably, chronic ablation of dopamine inputs disorganized these distributional representations in the striatum without interfering with mean value coding.
Two-photon calcium imaging and optogenetics revealed that the two major classes of striatal medium spiny neurons — D1 and D2 MSNs — contributed to this code by preferentially encoding the right and left tails of the reward distribution, respectively. We synthesize these findings into a new model of the striatum and mesolimbic dopamine that harnesses the opponency between D1 and D2 MSNs to reap the computational benefits of DRL.Medical Science
Molecular Basis for Differential V(D)J Recombination Mechanisms in Immunoglobulin Loci
No full textV(D)J recombination assembles variable region exons in antigen receptor loci and contributes to the exceptional diversity of immunoglobulins and T cell receptors. V(D)J recombination is initiated by the dedicated Recombination Activating Gene (RAG) endonuclease, which recognizes and cleaves at recombination signal sequences (RSSs) flanking each variable (V), diversity (D) and joining (J) gene segment. In the often multi-mega-base antigen receptor loci, how V(D)J recombination machinery coordinates with chromatin structural elements to ensure accurate and ordered rearrangements is an overarching question of the field. Over the past decade, studies focusing on the immunoglobulin heavy chain locus (Igh) demonstrated that cohesin loop extrusion-mediated RAG chromatin scanning is the underlying mechanism of strictly deletional rearrangements in Igh. However, how the immunoglobulin κ light chain locus (Igκ) enables both deletional and inversional Vκ-to-Jκ rearrangements remained, until recently, a mystery. This dissertation presents findings in mouse models and cell lines to support the model that orientation-independent Igκ primary recombination occurs through a short-range diffusion process dependent on the CTCF-based Cer/Sis element which separates the Vκ and Jκ domains. Importantly, strong Igκ RSSs are crucial for programming such diffusional Vκ-to-Jκ rearrangements, in contrast to the generally weaker Igh RSSs that rely on adjacent loop extrusion impediments for robust utilization during RAG scanning. After primary Vκ-to-Jκ joining-mediated removal of Cer/Sis, RAG linearly scans the newly proximal Vκ domain for predominantly deletional Igκ secondary recombination. However, unlike Igh, Vκ segments with strong RSSs can also be robustly utilized for inversional Igκ secondary recombination. In summary, the body of work presented in this dissertation elucidated the long-sought molecular mechanisms governing physiologically-important Igκ primary and secondary recombination. These findings also suggest that differential RSS strength can mediate distinct V(D)J recombination mechanisms in Igκ versus Igh and can potentially serve as an additional layer of regulation in other antigen receptor loci.Medical Science