University of Illinois Urbana-Champaign
Illinois Digital Environment for Access to Learning and Scholarship RepositoryNot a member yet
123813 research outputs found
Sort by
Mimicking nature with nanotechnology: engineering gold nanoparticles for targeted protein interactions
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Anita Wo, accepted the attached license on 2025-07-11 at 10:53.The student, Anita Wo, submitted this Dissertation for approval on 2025-07-11 at 15:01.This Dissertation was approved for publication on 2025-07-15 at 14:59.DSpace SAF Submission Ingestion Package generated from Vireo submission #22501 on 2025-10-21 at 10:05:51Colloidal nanoparticles show great versatility in biological applications, which include diagnostics, therapeutics, targeted drug delivery, biosensing, and bioimaging. Gold nanoparticles (AuNP) are particularly attractive materials due to several desirable properties including: 1) biocompatibility due to their physiologically inert nature, 2) size tunability within the range of 1-100nm, and 3) easy modification of surface chemistry composition. Control over the customizable properties of nanoparticle design allows for greater selectivity in its application. As the applications of nanoparticles are becoming increasingly integrated into healthcare, further research into the interactions of nanoparticles within complex living systems is required. Chapter 1 introduces the biomedical potential of nanoparticles and emphasizes the importance of understanding and predicting their behavior in biological environments. A central challenge in the field is that nanoparticles acquire a biological identity upon exposure to biological fluids due to biomolecular adsorption, which can lead to undesirable outcomes. Major translational barriers include biocompatibility, immune activation, and off-target effects, and reviews strategies for surface engineering to mitigate these issues. Fibrinogen, an abundant blood protein, was selected as a protein of interest to investigate the influence of nanoparticle surface design on protein binding behavior. In Chapter 2, a synthetic platform was developed in which dithiol peptides were grafted onto the surface of spherical AuNPs to mimic protein–ligand interactions. By tuning peptide density, it was possible to control conformational display on the nanoparticle surface, enabling selective binding to Mac-1 (αMβ2 integrin), a fibrinogen-binding receptor. The nanoparticles demonstrated strong binding affinity and specificity to Mac-1 at both the molecular and cellular levels, even outperforming native fibrinogen in some cases. Chapter 3 builds on these findings by introducing additional conformational constraint using cyclic peptides. A systematic library of cyclic peptide–AuNP conjugates was synthesized with spatial tuning. These constructs showed improved binding to Mac-1 at the molecular level, though further optimization is needed to translate similar binding outcomes at the cellular level. Together, this work provides a deeper understanding of how introducing peptide conformation on nanoparticle surfaces can result in protein mimics capable of targeted protein interactions
Forecasting without sequences: graph representations for dynamic systems in finance and beyond using GNNs
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Eamon Bracht, accepted the attached license on 2025-07-16 at 12:31.The student, Eamon Bracht, submitted this Dissertation for approval on 2025-07-16 at 12:38.This Dissertation was approved for publication on 2025-07-17 at 18:30.DSpace SAF Submission Ingestion Package generated from Vireo submission #22602 on 2025-10-21 at 10:06:09Forecasting complex systems has traditionally been framed as a sequence learning problem, assuming that past trajectories contain sufficient information to predict future states. However, in many real-world domains—including financial markets—this assumption fails due to stochasticity, structural shifts, and hidden dynamics. This dissertation proposes a conceptual shift: to view forecasting not as extrapolating sequences, but as learning from structure. The work is divided into three major parts. First, we introduce the Sparse Spatio-Temporal Neural Network (sSTNN), a scalable architecture for large-domain forecasting tasks. While sSTNN performs well on smooth systems like ocean temperature, its failure on deforestation forecasting tasks exposes the limits of sequence-based approaches when faced with abrupt, structurally-driven dynamics. Second, focusing on financial markets, we demonstrate that meaningful structural information persists even at extremely short intraday timescales. Using advanced community detection methods on financial correlation networks, we show that sectoral organization can be reliably recovered from minutes or hours of return data. These findings suggest that market behavior is better captured through evolving relational structures rather than isolated time series. Finally, building on these insights, we develop a predictive framework using Message Passing Graph Neural Networks (MPGNNs) trained on financial graphs. Our models, incorporating both node and edge features, forecast future index movements with statistically significant accuracy across multiple temporal resolutions. Crucially, the success of these graph-based models is not solely explained by conventional sector clustering, indicating that deeper latent topological signals underlie market dynamics. Together, these contributions argue for a structural paradigm in forecasting complex systems, where topology and relational information serve as the primary modeling substrate. Beyond financial markets, the methods and insights developed here lay the groundwork for applying graph-based forecasting to a broader range of dynamic, structurally complex domains
Improving integrated pest management for plum curculio in Illinois peaches
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Karuna Kafle, accepted the attached license on 2025-07-17 at 10:48.The student, Karuna Kafle, submitted this Thesis for approval on 2025-07-17 at 11:01.This Thesis was approved for publication on 2025-07-21 at 14:58.DSpace SAF Submission Ingestion Package generated from Vireo submission #22630 on 2025-10-21 at 10:06:11Plum curculio (Conotrachelus nenuphar) is a native pest and internal feeder and is a major threat to peaches and apples in Illinois. Orchard surveys were conducted in 2023 and 2024 to assess the effectiveness of different traps and lures for monitoring and capturing adult plum curculios in commercial orchards in west-central and southern Illinois. In 2023, plum curculio adults were successfully collected from Orchards E and W earlier in the season using traps baited with a benzaldehyde lure. However, adults collected using the lures declined later in the season despite their continued presence in the field, possibly indicating reduced attractiveness of the lure. The captured adults were counted and brought to the laboratory for further studies. In 2024, no adults were captured from orchards due to an early season freezing event, which led to a lack of peaches in Orchards E and W. A laboratory experiment using a Y- tube olfactometer was conducted to test if the lure is attractive enough to plum curculios when suitable host fruit (peach and apple) is present. In Y tube, we tested the responses of plum curculio adults to Benzaldehyde + Grandisoic acid (BZ+GA) and/or Methyl Salicylate + Grandisoic acid (MeSA+GA) versus host fruits (apple and/or peach). The results revealed no significant difference in the attractiveness of the lure when the host (either apple or peach) was present. This finding explains why the lure’s attractiveness declined as the season progressed. We used molecular approaches to identify the strain of plum curculio in Illinois. The adults collected from the field were used for DNA extraction and PCR was performed using Wolbachia specific primers (wsp) to identify associated Wolbachia strains. Previous studies reported that different strains of plum curculio are infected with distinct Wolbachia strains; wcne3 infects the bivoltine southern strain, while wcne1 and wcne2 infect the univoltine northern strain. Our study revealed that most adults collected from Orchard E and W were the southern strain, with only a single individual identified as the northern strain. This suggests that Illinois hosts both northern and southern strains, indicating a possible northward expansion of the southern strain’s range from the original boundary. The findings of this study highlight the limitations of commercial lures in mid-summer and provide evidence of an additional generation of plum curculio in Illinois. The findings indicate the need for refined IPM practices, including aligning insecticide application timing with the insect’s life cycle to improve plum curculio management in Illinois
The law and psychology of altruism in allocating scarce medical resources
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Qiaoyuan Zhi, accepted the attached license on 2025-07-17 at 14:56.The student, Qiaoyuan Zhi, submitted this Dissertation for approval on 2025-07-17 at 15:11.This Dissertation was approved for publication on 2025-07-18 at 06:41.DSpace SAF Submission Ingestion Package generated from Vireo submission #22641 on 2025-10-21 at 10:06:11This experimental study (N = 2706) investigated the relationship between costly altruism in medical decision-making and various predictors, such as health status and demographic characteristics. It also examined bias in social expectations about costly altruism, and assesses the impact of inducing individuals to be more other-regarding. We found that participants tended to expect others to use risk-based criteria, prioritizing individuals at higher medical risk. They did not expect others to rely on remaining life expectancy, quality-adjusted life years (QALYs), or identity-based biased approach such as gender or race. These expectations largely aligned with actual allocation patterns, though there were some differences in degree. For instance, participants did not adopt identity-based stereotypes of altruism, expecting people with greater altruistic reputations to be more altruistic. In practice, these groups did not give significantly more resources than others, and older adults even gave less. Additionally, participants correctly anticipated no racial bias against Black recipients, but underestimated the extent to which Black individuals would actually be favored: participants gave more resources to Black recipients than to White recipients. As to factors designed to increase other-regardingness, the evidence failed to show that the identifiability of a target generally enhances altruism, and it may even reduce it among certain groups. The study adds to the existing literature on expectations about altruism and altruistic behavior, expanding it to consider circumstances that involve a higher level of self-sacrifice than donating, volunteering, or allocating a fixed amount of money in dictator games. The ambition of this study is to provide empirical evidence to help lawmakers think about the fair allocation of scarce medical resources not just during a pandemic, but even during normal times
Measurement limits and performance analysis of multilayer insulation under vacuum and cryogenic temperatures using a modified calorimeter-bar setup
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Saptarshi Joshi, accepted the attached license on 2025-07-21 at 19:43.The student, Saptarshi Joshi, submitted this Thesis for approval on 2025-07-21 at 19:56.This Thesis was approved for publication on 2025-07-23 at 15:00.DSpace SAF Submission Ingestion Package generated from Vireo submission #22684 on 2025-10-21 at 10:06:18High-performance cryogenic electric machines, such as superconducting motors, require exceptional thermal insulation to minimize heat leakage and maintain operational temperatures below 50 K. Multilayer insulation (MLI) is widely used in such systems due to its low thermal conductance, but its performance is highly sensitive to contact pressure, vacuum level, and structural configuration—making accurate experimental characterization essential. This thesis presents the design, development, and implementation of a modified ASTM D5470-based one-dimensional (1D) calorimeter bar apparatus to measure the through-thickness thermal conductance of MLI under vacuum at both ambient and cryogenic temperatures. The apparatus integrates mirror-polished metal bars, high-precision temperature sensing, and a high-vacuum chamber to enable controlled steady-state heat flow through MLI samples. Cryogenic conditions were achieved using both liquid nitrogen and a cryocooler-thermal strap system. Experimental results at ambient temperatures showed that stainless steel bars provided improved measurement reliability due to steeper axial temperature gradients. Tests at cryogenic temperatures demonstrated clear trends in conductance with respect to layer count, compression, and shielding, but failed to reach the ultra-low conductance values theoretically expected of high-performance MLI. Analytical modeling and uncertainty analysis revealed that lateral radiative losses and shallow temperature gradients led to significant overestimation of true conductance at low flux levels. The results indicate that while the adapted calorimeter bar method is suitable for moderate conductance characterization, its current configuration imposes a practical resolution limit near 1 W/m²·K. This study establishes both the capabilities and limitations of the ASTM D5470 method for cryogenic MLI evaluation and informs future designs for ultra-low conductance measurement systems
Information-Theoretic Outer-Bounds from Sphere-Packing, Compactness, and Zero-One Laws
Get PDFInformation-theoretic outer-bounds are derived using sphere-packing arguments for the point-to-point channel and the relay channel. The arguments rely on novel applications of compactness and zero-one exponent laws to address issues of coordination and distortions to the codeword orbits caused by generic relay encoding functions. The capacity of the relay channel is I(Xs; Ŷr, Yd|Xr) for I(Ŷr; Yr|xr) H(Yr|Ŷr,Xr) and a coarse compression regime where H(Yr|Xs,Xr) ≤ H(Yr|Ŷr,Xr). Both regimes support p(xs)p(xr)p(ŷr|yr, xr)p(yr, yd|xs, xr) which corresponds to simple compress-forward (no Wyner-Ziv). The coarse compression regime also supports p(xs, xr)p(ŷr|xr)p(yr, yd|xs, xr) which corresponds to partial decode-forward
Factors Affecting Migratory Bird-Window Collisions and Effective Mitigation Strategies
No full textBird-window collisions pose a major threat to bird populations, killing up to a billion birds annually in the United States alone. Migratory birds, many of which use the stars and Earth's magnetic field to navigate during the night, are particularly vulnerable to collisions since anthropogenic factors such as light pollution can disrupt their ability to navigate. Factors such as species behavior, surrounding vegetation, and window design have been found to influence collision rates. Research has also identified several effective mitigation strategies, including bird-safe window treatments, reduced artificial light at night, and public education. Addressing the threat of window collisions is crucial to protect avian biodiversity worldwide and preserve the ecological benefits provided by birds
Biological Neural Networks as the Forefront of AI Processing
No full textGraphic processing units (GPUs) are a major component of artificial intelligence (AI) processing power. As AI becomes more sophisticated and more processing power is needed to run these intellectual models, a growing concern of energy demand and extensive AI training becomes an increasing concern. To create more sophisticated machine learning algorithms, scientists in the field of neuromorphic computing studied the brain for its ability to efficiently process and store information. Finding a way to incorporate the brain directly into computing may create novel algorithms to meet the increasing demands of information processing
Pathophysiology of Postpartum Depression: Etiology and Interplay of Structural and Functional Brain Changes
No full textPostpartum depression (PPD) affects a significant portion of new mothers, leading to severe disruptions in maternal mental health, such as persistent feelings of sadness, anxiety and emotional numbness. These symptoms not only hinder the mother’s well-being but also interfere with critical maternal-infant bonding and early caregiving, which can have lasting developmental consequences for the child. Despite the well-documented emotional and cognitive consequences of PPD, the neurobiological mechanisms underlying this condition remain are still not fully understood. Structural and functional brain alterations in areas such as the prefrontal cortex (PFC), hippocampus, and amygdala have been implicated in the development of PPD. Neuroimaging studies offer promising insights into the brain changes associated with this mood disorder. Understanding these modifications could pave the way for earlier identification and more targeted interventions to improve maternal mental health outcomes
THE MUSICAL AESTHETICS OF JEAJOON RYU: AN INTERPRETIVE STUDY OF SUITE PER PIANOFORTE NO. 2
No full textThis dissertation introduces the Korean composer Jeajoon Ryu (b. 1970) to an international audience through a comprehensive analysis of his Suite per Pianoforte No. 2
(2023). It argues that Ryu’s compositional voice emerges from a synthesis between reverence for the Western classical tradition and engagement with contemporary social realities. Through this synthesis, Ryu forges a unique individual aesthetic.
The Suite per Pianoforte No. 2—derived from Ryu’s two-person opera and song cycle Apartment (2021), a work exploring the psychological and social narratives connected to modern urban living spaces—serves as the paradigmatic work for exploring this fusion. The movements for piano solo in Apartment are uniformly titled "Prelude" and vocal movements have programmatic titles. In the suite, however, Ryu recontextualizes each movement to reflect specific genres, forms, or techniques of the Western classical instrumental works that serve as its models. This process, achieved through the integration of baritone melody and piano accompaniment into the solo piano writing, not only increases the technical and expressive demands on the performer but also generates a new structural coherence.
Through the development of interrelated rhythmic, intervallic, and harmonic ideas, each movement conveys its own narratives and emotions centered on the theme of human living spaces. Detailed analysis reveals that these shared materials create a strong underlying sense of
unity across the diverse movements. Ultimately, this study illuminates how Ryu harmonizes musical tradition with contemporary topical concerns, positioning his Suite per Pianoforte No. 2 as a notable contribution to the 21st-century piano suite repertoire