150813 research outputs found
Sort by
Development of Dual Extruder Biomaterial 3D Printer
No full textThis research presents the design and fabrication of a novel dual-extruder biotic 3D printer for the precise deposition of natural biocomposites using organic materials such as pectin, chitosan, and cellulose. Unlike traditional FDM printers that rely on thermoplastic extrusion, this printer employs a syringe-based mechanical extruder capable of depositing viscous biomaterial hydrogels. The integration of a first-of-its-kind dual-extruder system enables the fabrication of multi-material prints and the exploration of biomaterial composites and complex geometric structures, thereby advancing sustainable, bio-inspired manufacturing.
This thesis emphasizes the machine engineering aspects of the printer's development, including project motivation, systematic design methodology, component design and fabrication, testing, and exploration of future work. Notable features of the system include user-friendly operation for non-experts, open-source accessibility, and compatibility with a wide range of biomaterials. By addressing existing limitations in biomaterial 3D printing technology, this work provides a robust platform to support future research in biomaterials, sustainable additive manufacturing, and bio-inspired design. Furthermore, the open-source nature of the printer fosters innovation and collaboration, accelerating the adoption of sustainable materials and manufacturing methods.S.M
On the Acquisition of Formal Semantics in Statistical Models of Language
No full textThe increasingly impressive performance of recent large language models raises a crucial question: to what extent can such models, trained solely on text, develop an understanding of language grounded in the semantics of the underlying domain? Progress on this question carries significant practical and philosophical implications for the relationship between meaning, understanding, and the capacity to exhibit seemingly intelligent behavior.
This thesis makes two primary contributions. First, it develops a scientifically rigorous approach to studying what statistical models of language can understand about language based on the formal semantics of programming languages. Specifically, it leverages the probing classifiers framework: training small classifiers to find encodings of program semantics within the model's internal representations. A main insight is that the clean separation between syntax and semantics in this domain allows for greater control in experimental design. It introduces two new techniques. The first, semantic probing interventions, is a general methodology for distinguishing whether the probe's measurements reflect (1) the learned representations of the language model encode semantics or (2) that the probe itself has learned to infer semantics from representations of pure syntax. The second, latent causal probing, is a formal framework for probing that provides a robust empirical methodology for studying whether language models are able to access the latent concepts that underlie the text they observe during training. A key innovation is to create a single structural causal model that unifies (1) the data generation process underlying the text used to train the language model and (2) the steps of a probing experiment. This makes it possible to conduct a causal analysis that intervenes on the data generation process to trace the influence of the latent variables in the training data through the model's internal representations.
The second core contribution of this thesis consists of a series of experimental studies. Specifically, we train a language model on a synthetic grid-world navigation task, then probe the model's learned representations for encodings of the unobserved, intermediate world states. By leveraging the techniques we develop, the results deliver strong empirical evidence that statistical models of language are latent concept learners: capable of inducing the latent variables that underlie the generation of their training data, despite being trained only to model a conditional distribution over tokens.Ph.D
Characterizing Variation in Healthcare across Time and Providers using Machine Learning
No full textModeling healthcare decisions and their outcomes is a complex problem. In addition to being affected by patient characteristics, the prognosis can vary depending on when the patient is receiving care, and treatment decisions can vary depending on who makes the decisions. In this thesis, we consider two axes of variation in healthcare: over time and across providers. For both axes, we focus on identifying when variation exists, characterizing the patients who are affected by such variation, and addressing shifts due to this variation. The solutions we propose draw ideas from causality and dataset shift.
In the first part of this thesis, we address these three aspects for variation over time. First, we create an algorithm that can detect when a model is affected by change over time and identify sub-populations where the model is more affected. We use our algorithm to perform a large-scale study of temporal shifts in health insurance claims. We demonstrate changes over time are prevalent in healthcare and examine case studies to better understand the drivers of such changes. Next, we examine how to learn a model that can perform well on current data. As data from the current time period is limited, we consider several methods that can leverage sequences of historical data to learn a good image classification model for the final time step. We build a benchmark for evaluating these methods on sequences constructed from synthetic shifts and validate our conclusions on a real-world dataset.
In the second part of this thesis, we address similar questions for variation across providers. First, we create a statistical approach to test whether significant variation exists across providers. Our approach involves learning a model of treatment decisions with provider-specific random effects. We perform a case study on first-line type 2 diabetes treatment and find significant variation exists across providers. Then, we develop an algorithm for identifying regions of patients with the most disagreement between providers. We formalize this as a causal inference problem, where disagreement is defined by the causal effect of the provider on the treatment decision. We illustrate this algorithm on first-line type 2 diabetes and Parkinson's treatment decisions and uncover regions of variation that align with uncertainty in clinical guidelines.
In the third part of this thesis, we build a tool for examining the effects of variation over time or across providers for individual patients. We use a large language model built on electronic health record concepts to generate patient trajectories. To enable interventions on time and provider, we introduce new tokenizations for these concepts. We also incorporate a structural causal model for patient visits to allow for generation of interventional and counterfactual trajectories. We hope the model in this part of the thesis can be used to answer additional questions about how patient trajectories would change if they were treated during a different time period or by a different provider.Ph.D
Variations in approaches to urban climate adaptation: Experiences and experimentation from the global South
No full textIn recent years, an increasing number of local governments are recognizing the impact of climate change on different urban sectors. This has led many to pursue climate adaptation planning, seeking to achieve preparedness through reducing vulnerability and enhancing resilience of populations, assets, and municipal operations. Although cities typically share these common goals, many are electing to pursue different planning approaches. In this paper, we examine three climate adaptation planning approaches in the cities of Quito (Ecuador), Surat (India), and Durban (South Africa) and analyze the trade-offs associated with different planning pathways and different forms of stakeholder involvement. We assess the potentials and limitations of these different approaches, including their implications for enhancing government integration and coordination, promoting participation and adaptive capacity of vulnerable groups, and facilitating overall urban resilience. We find that, in order to gain widespread commitment on adaptation, sustained political leadership from the top, departmental engagement, and continued involvement from a variety of stakeholders are integral to effective decision-making and institutionalization of programs in the long run. When climate adaptation is advanced with a focus on learning, awareness, and capacity building, the process will likely lead to more sustained, legitimate, and comprehensive adaptation plans and policies that enhance the resilience of the most affected urban areas and residents
Dynamic Markers
No full textWhen I was a child, I was certain that all clouds came from New Jersey. After passing through the Lincoln Tunnel, I-95 would gradually ascend, lifting our car to eye level with the billowing clouds emerging from beneath us. These clouds rose from the Meadowlands, a great marsh just two miles west of Manhattan, a landscape that has become defined by the infrastructure that occupies it. Nearly equal in land mass and opportunity to Manhattan, this landscape managed to resist holistic transformation due to our inability to control its water. Rather than becoming a prosperous site for agriculture in the 19th century, or the next metropolis in the early 20th, the Meadowlands fell out of focus and became a site to absorb the infrastructural networks needed to uphold rapid development at its edges.
The Meadowlands was sutured shut by the networks interlaced through it in an attempt to erase the failures of the past. Utilizing this landscape as an urban sponge neglected that the marsh hosted a series of ecological infrastructures of its own. The Meadowlands' soft, uncertain ground once managed variations in the water level, but the draining of the ground that came with development reduced its capacity, making pump stations essential for managing water in inhabited areas. Unlike the other forms of infrastructure in the Meadowlands, the presence of the pump station is subdued, its invisibility upholds the illusion that the developments within this landscape are not threatened by their surroundings. However, steady sea level rise and an increase in storm surges have caused these pumps to fail, pulling the veil on their existence and more importantly, the essential role they play in our continued occupation of this landscape. The urgent need to increase the capacity of the pump station provides an opportunity to reconsider their agenda.
This thesis proposes the Dynamic Marker, a new type of infrastructure that redefines the relationship between human systems and ecological flows. Grafted onto existing pump stations in the Meadowlands, it releases water as mist from 800 feet in the air, transforming the hidden mechanics of water management into a moment of wonder. The Dynamic Marker fosters microclimates and ecological connections, transforming infrastructure into a dynamic process that evolves with its surroundings. Over time, it becomes both a memorial to the marsh and a provocation for the future, inviting a rethinking of infrastructure as a participatory and adaptive force that responds to its surrounding ecology.M.Arch
Brain Markers of Resilience to Psychosis in High-Risk Individuals: A Systematic Review and Label-Based Meta-Analysis of Multimodal MRI Studies
No full textBackground/Objectives: Most individuals who have a familial or clinical risk of developing psychosis remain free from psychopathology. Identifying neural markers of resilience in these at-risk individuals may help clarify underlying mechanisms and yield novel targets for early intervention. However, in contrast to studies on risk biomarkers, studies on neural markers of resilience to psychosis are scarce. The current study aimed to identify potential brain markers of resilience to psychosis. Methods: A systematic review of the literature yielded a total of 43 MRI studies that reported resilience-associated brain changes in individuals with an elevated risk for psychosis. Label-based meta-analysis was used to synthesize findings across MRI modalities. Results: Resilience-associated brain changes were significantly overreported in the default mode and language network, and among highly connected and central brain regions. Conclusions: These findings suggest that the DMN and language-associated areas and central brain hubs may be hotspots for resilience-associated brain changes. These neural systems are thus of key interest as targets of inquiry and, possibly, intervention in at-risk populations
American (Ise): On the Lifecycle of Stadiums in the United States
No full textWhen the Kingdome in Seattle was completed in 1976, it was celebrated as a marvel of modern engineering, expected to last for centuries. Yet, in an ironic twist, it was demolished by implosion in 2000, surviving only twenty-four years. The Kingdome epitomizes the issue of short lifespans that has plagued American stadiums since the post-war era. A broad survey of these structures reveals an average lifespan of just three decades—a startlingly brief tenure for buildings of their scale and significance. These stadiums also follow a distinctive model of renewal. Similar to the Shikinen Sengu ritual at the Ise Shrine, a new stadium is often constructed adjacent to its predecessor. However, unlike Ise, where materials from the old shrine are reused and disseminated throughout Japan’s network of shrines, old stadiums are almost always demolished and discarded. This thesis seeks to superimpose Ise as a model onto American stadiums, envisioning an architecture that embraces both impermanence and longevity through circularity. Investigations into the barriers to circularity specific to stadiums serve as the foundation for design proposals, spanning scales from the detail to the site. The project ultimately imagines a stadium in a constant process of disassembly and renewal, where its spatial and programmatic potential challenge paradigms of completeness. In the context of a climate crisis demanding waste reduction, and for a typology notorious for its excess, stadiums can learn to do more with less.M.Arch
Coarse Modality
No full textOne of the early successes of the application of possible worlds semantics to the analysis of natural language is Kratzer’s account of modality. A large part of the subsequent literature on modals has sought to expand the crosslinguistic coverage of that framework, and, in so doing, many new generalizations and constraints have been proposed and re-examined. The present dissertation situates itself within this tradition and makes both an empirical and theoretical contribution. Using the Italian adverb magari as the main empirical source, it will be argued that there exists a previously unnoticed type of modality which is referred to here as “coarse”. Its most evident manifestation is a special type of epistemic possibility, one that comes with an “antievidential” requirement. Antievidential possibility in assertions and questions is discussed in Chapters 1 and 3 respectively. Chapter 2 frames coarse modality as a more general phenomenon that comes about through modification of modal expressions. The theoretical argument of this dissertation is a novel corroboration of Kratzer’s premise semantics approach. It will be argued that the most natural and general account of coarse modality is possible by utilizing the premise set, a powerful resource of the system, in a novel way.Ph.D
Beyond the Bloom: Invasive Seaweed Sargassum spp. as a Catalyst for Sustainable Agriculture and Blue Economy—A Multifaceted Approach to Biodegradable Films, Biostimulants, and Carbon Mitigation
No full textThe Anthropocene has ushered in unprecedented environmental challenges, with invasive seaweed blooms emerging as a critical yet understudied facet of climate change. These blooms, driven by nutrient runoff and oceanic alterations, disrupt ecosystems, threaten biodiversity, and impose economic and public health burdens on coastal communities. However, invasive seaweeds also present an opportunity as a sustainable resource. This study explores the valorization of Sargassum spp. for agricultural applications, focusing on the development of biodegradable bioplastics and biostimulants. Field trials demonstrated the effectiveness of Marine Symbiotic® Sargassum-derived biostimulant in distinct agricultural contexts. In the Dominican Republic, trials on pepper crops showed significant improvements, including a 33.26% increase in fruit weight, a 21.94% rise in fruit set percentage, a 45% higher yield under high-stress conditions, and a 48.42% reduction in fruit rejection compared to control. In Colombia, trials across four leafy green varieties revealed biomass increases of up to 360%, a 50% reduction in synthetic input dependency, and enhanced crop coloration, improving marketability. Additionally, Sargassum-based biofilms exhibited favorable mechanical properties and biodegradability, offering a sustainable alternative to conventional agricultural plastics. Carbon credit quantification revealed that valorizing Sargassum could prevent up to 89,670 tons of CO2-equivalent emissions annually using just one Littoral Collection Module® harvesting system, while biostimulant application enhanced carbon sequestration in crops. These findings underscore the potential of invasive seaweed valorization to address multiple climate challenges, from reducing plastic pollution and GHG emissions to enhancing agricultural resilience, thereby contributing to a sustainable Blue Economy and aligning with global sustainability goals
Synthesis and oxidation behavior of Cr alloyed uranium borides at high temperatures
No full textFollowing the nuclear accident at Fukushima Daiichi Power Station in 2011, an urgent need for safer, more economical, and versatile nuclear fuels has arisen. In recent years, uranium boride (as a tetraboride and diboride) has been further investigated as a candidate fuel form for its high thermal conductivity, high melting point, high uranium loading, and potential for dual use as a fuel and burnable absorber. In this work, the synthesis, structural behavior, and oxidation behavior of uranium borides and chromium- and yttrium- alloyed uranium borides are investigated. The structure of the synthesized uranium borides and chromium- and yttrium- alloyed uranium borides were probed using synchrotron X- ray Powder Diffraction (XRD) and Pair Distribution Function (PDF) analysis with in-situ heating. The methods and challenges in synthesizing uranium boride and chromium- and yttrium-alloyed uranium boride, as well as the consequential thermophysical and oxidation properties of these potential fuel forms, are elucidated in this work.S.B.S.M