21794 research outputs found
Sort by
Operationalizing Care In A Jesuit Catholic Urban High School
No full textBlack males in urban schools typically perform below their White peers academically, leading educational researchers to seek effective instructional activities to help close this achievement gap. One way to improve the academic attitudes and behaviors of Black males is to strengthen the relationships between teachers and students through the Ignatian concept of care applied in Jesuit schools, as relationships play a pivotal role in teaching and learning and in the application of culturally relevant and Ignatian pedagogies. This qualitative study, through interviews of teachers and students in one urban area Jesuit Catholic high school with an all-male, predominately Black population, explores how care is expressed by teachers and experienced by students and its impact on academic attitudes and behaviors of the students. The findings of the study suggest that care is present in the classrooms of the school and infuses all corners of the educational relationship triangle. The findings also imply that the teachers are operationalizing elements of culturally relevant pedagogy
New Generation Of Machine Learning Models To Improve Prediction And Optimization In Energy Systems
No full textEnergy systems are experiencing a significant transformation marked by emerging requirements associated with the integration of renewables, electric vehicles, and distributed generation into the modern power grid. These changing dynamics and requirements mandate a new set of (i) predictive methods that use streaming industrial data to improve situational awareness and (ii) prescriptive models to offer improved control and optimization of operational decisions. An underlying enabler for these new set of methods is the availability of data, processing capabilities, and the innovative use of machine learning (ML) methods to address domain specific deployment challenges. This dissertation develops new approaches that revisit and transform two fundamental problems in energy systems: 1) federated prognostics: leveraging ML for privacy-preserving predictive models for predicting remaining life of lithium ion batteries, 2) ML-enhanced operations optimization: incorporating ML techniques for prescriptive models to accelerate solution methods for a central operational problem in power systems called stochastic unit-commitment (SUC).
The first problem concentrates on predicting the remaining useful lifetime of lithium-ion batteries, utilizing a federated learning (FL) approach. The proposed approach offers a paradigm shift from the existing methods that build on centralized data collection and processing from various clients, which suffer from communication and processing bottlenecks due to the substantial volume of information. Additionally, these methods raise significant privacy concerns and potential liabilities related to data breaches. To address these challenges and to enable scalable deployment of battery management systems, we propose a novel approach: a federated battery prognosis model. This model offers prognostics methods that decentralize the processing of battery standard data, such as current-voltage-time-usage information, with a focus on prioritizing privacy. Rather than exchanging raw data, our framework only shares model parameters with the central server, reducing the load on communication channels and safeguarding data confidentiality.
In the second research problem, we address computational scalability of SUC, a computationally challenging problem with complex constraints such as transmission line capacities and ramping limits. The SUC problem determines the most efficient combination of generating units for commitment and identifies optimal generation levels, by considering a large number of constraints and a significant uncertainty from demand fluctuations. Our research aims to develop new methods that use ML to enhance existing solution methodologies to reduce computation times. Specifically, our study introduces a hybrid approach that combines reinforcement learning and optimization for improving Benders decomposition implementation in SUC problems
Hijacking Large Language Models Via Adversarial Incontext Learning
No full textIn-context learning (ICL) has emerged as a powerful paradigm leveraging LLMs for specific downstream tasks by utilizing labeled examples as demonstrations in the precondition prompts. Despite its promising performance, ICL suffers from instability with the choice and arrangement of examples. Additionally, crafted adversarial attacks pose a notable threat tothe robustness of ICL. However, existing attacks are either easy to detect, rely on external models, or lack specificity towards ICL. This thesis introduces a novel transferable attack for ICL to address these issues, aiming to hijack LLMs to generate the targeted response. The proposed hijacking attack leverages a gradient-based prompt search method to learn and append imperceptible adversarial suffixes to the in-context demonstrations. Extensive experimental results on various tasks and datasets demonstrate the effectiveness of our hijacking attack, resulting in distracted attention towards adversarial tokens and consequently leading to unwanted target outputs. We also propose a defense strategy against hijacking attacks through the use of extra demonstrations, which enhances the robustness of LLMs during ICL. Broadly, this work reveals the significant security vulnerabilities of LLMs and emphasizes the necessity for in-depth studies on the robustness of LLMs related to ICL
Dietary Folate Regulates Organismal Growth And Longevity
Get PDFFolate (Vitamin B9) is an essential cofactor in one-carbon transfer reactions. It is involved in multiple critical processes including DNA synthesis, DNA methylation, protein synthesis and antioxidant defense mechanisms. Inadequate intake and availability of folate is a known risk factor for several diseases. The effect of folate deficiency in causing adverse health outcome is well documented. Recently, excess folate exposure has been associated with deleterious effects on health in humans and animal models. However, the underlying mechanisms involved in mediating such effects are poorly understood, requiring extensive investigations. This work was conducted to understand the effect of excess folate on health and lifespan in mice and C. elegans with an emphasis on investigating the role of the microbiome in mediating folate’s effects. Reduction of folate production in the gut using the antibiotic SST resulted in distinct bacterial clustering at various taxa levels, downregulation of mTOR signaling and an upregulation of antioxidant genes, suggesting a direct link between folate and aging. To further elucidate the mechanistic effects of folates on lifespan and health span, we used the well characterized C. elegans aging model. Excess folate resulted in the reduction of median lifespan and oxidative stress resistance in wild type C. elegans. Furthermore, NAC administration failed to rescue oxidative stress under excess folate suggesting that cytosolic oxidative stress is not the main driver of excess folate induced aging phenotype. Our results suggest that dietary folate availability is an important conserved regulator of lifespan. The results from this study are critical to human health because consumption of excess FA through fortification and supplementation can potentially impact health and lifespan in humans
Generalized Coderivative-Based Newtonian Methods In Nonsmooth Nonconvex Optimization And Applications
No full textThis dissertation proposes and develops new Newton-type methods to solve nonconvex and nonsmooth optimization problems with justifying their fast local and global convergence by means of advanced tools of variational analysis and generalized differentiation. The objective functions belong to a broad class of prox-regular functions with specification to constrained optimization of convex and nonconvex structured sums. The proposed algorithms are formulated in terms of the second-order subdifferential of such functions that enjoy extensive calculus rules and can be efficiently computed for broad classes of extended-real-valued functions. Further applications and numerical experiments are conducted for the Lassoproblems, the box constrained problems of quadratic programming, and the fast best subset selection problems, which play a crucial role in statistics and machine learning
Verifying And Analyzing Ultra-Compact X-Ray Binaries Via Reflection Spectroscopy
Get PDFA low-mass X-ray binary (LMXB) is a compact object, such as a neutron star (NS) or black hole (BH), which is accreting material from a stellar companion via a process called Roche-lobe overflow. An ultra-compact X-ray binary (UCXB) is a subset of these systems defined by a shorter orbital period (\u3c 80 minutes, compared to the hours to days seen in LMXBs). This shorter orbital period implies that the companion is more compact than a main sequence type star, often being a white dwarf (WD). These systems are important to study, as the compact nature of both objects in a tight orbit produces low frequency gravitational waves that will affect the next generation of multi-messenger astronomy. By observing UCXBs we can help to understand every facet of these unique systems before such missions begin. Over the course of this dissertation, we study a NS-WD UCXB called 4U 0614+091 and a NS UCXB candidate called SLX 1735-269. The primary method used to study these systems is a process called reflection modeling. This process involves collecting X-ray spectra, and modeling various components; a thermal contribution from the NS, non-thermal contribution from an X-ray corona surrounding the compact object, and sometimes thermal contributions from the accretion disk itself when necessary. These make up the standard continuum for an LMXB, but an additional component is applied which accounts for photons from the corona illuminating the disk and being reprocessed by the elements therein. These elements differ between UCXBs and LMXBs, as UCXBs tend to have an overabundance of oxygen with respect to solar, while being nearly devoid of hydrogen, resulting in a unique feature in the low-energy band. We use this fact to attempt to verify the UCXB nature of SLX 1735-269 by comparing reflection models applied to the data. These models contain different chemical abundances, with some designed for standard LMXBs with a main sequence companion and another designed specifically for UCXBs with a WD companion. We find that it is more likely that the source is ultra-compact in nature, but our timing analysis yields no orbital period, so we can not verify this beyond the indirect evidence of abundances. We also use this method to analyzea series of observations of 4U0614+091, which is a confirmed UCXB. This source varies in flux quasi-periodically over the course of a few days. By capturing the source at various stages along this flux evolution and modeling the reflection, we find that the inner disk appears to move away from the NS at the lowest flux state. This is analogous to what is seen in BH LMXB systems, and explains an apparent anti-correlation in the flux of the illuminating corona and that of the reflected emission, as the accretion disk is physically further from the illuminating source. Finally, we focus on a long term analysis of the same source, taking 51 archival spectra and modeling them with a UCXB reflection abundances. We find the source in different states throughout the course of these observations, which allows us to study how certain reflection parameters change as the source varies overall. We find that our results are consistent with the X-ray corona moving closer to the NS during high flux states, where the bulk of the emission is at low energies. This unique subset of sources can help us understand accretion at various scales by drawing comparisons to more massive black hole counterparts
Ultrafast Imaging Of Electrons And Ions In Strong-Field Ionization Using Few-Cycle Laser Pulses
No full textAll chemical, biological, and physical reactions are driven by the correlated motions of electrons and their interaction with their parent ions. Probing and capturing the dynamics of the correlations of electrons and ions will reveal the details of the fundamental mechanism at which certain reactions occur which tends to be difficult to probe in the world of traditional chemistry. This will also provide chemists with deeper insights on how to control and optimize chemical reactions. However, the dynamics of electrons and ions happen at an ultrashort timescale within the femtosecond and attosecond time range. Several techniques have been developed in the past to capture these dynamics in their natural timescale and one of such techniques is the 3D velocity map imaging.This thesis describes in detail the development, implementation, and validation of the 3D velocity map imaging technique for the imaging of the momentum distribution of electrons and ions produced from the strong field ionization of atoms and molecules using an ultrafast laser system. An all-optical imaging technique that employs the use of a fast scintillator screen and a SiPMT was implemented for the 3D imaging of electrons. With this technique, we were able to detect two electrons in coincidence with the shortest possible deadtime of ~0.48ns. We also developed a new two-camera imaging system which serves as an alternative and cost-effective method for performing 3D momentum imaging of particles without the need for specialized timing equipment. This new technique can detect and resolve two ions with different masses and can achieve a temporal resolution of 2 ns when detecting photoelectrons. It can also be operated at a higher count which makes it a suitable approach to perform covariance imaging experiments. While the 3D momentum imaging technique is a powerful tool for probing electron and ion dynamics, the carrier-envelope phase (CEP) is also an important parameter to consider when probing electron dynamics using few-cycle laser pulses. Previously, the Li group has developed a way for directly measuring the CEP using the attoclock (angular streaking) technique. In this thesis, we demonstrate for the first time how we can exploit the CEP effect to measure the deflection angle of electrons for circularly polarized light using the phase-resolved attoclock technique. With this technique, we were able to investigate the nonadiabatic motion of electrons under the tunneling barrier and the effect of the coulomb potential on the deflection angle of the electron. Our result reveals that the coulomb potential dominates the dynamic motion of the electrons while nonadiabaticity and tunneling delay plays a minor role
Investigation Of Inhibition Of Ferroptosis Process By Small Molecules Having Antioxidant Properties
No full textInvestigation of Inhibition of ferroptosis process by small molecules having antioxidant propertiesBy Vibha Deshpande July 2024
Advisor: Dr. Aloke DuttaMajor: Pharmaceutical Sciences Degree: Master of Science Introduction/ Objective Ferroptosis is a caspase-independent form of regulated cell death driven by iron-induced formation of lipid peroxides that accumulate at toxic levels causing cell death. Under physiological conditions, the cell combats lipid peroxidation with the help of selenoprotein GPX4. It has a central role in the ferroptosis pathway as it reduces lipid peroxide to lipid alcohol. Our study uses the utility of employing the ferroptosis-inducing compound RSL3 in cellular models to study the mechanism of production and inhibition of ferroptosis. Specifically, our investigation aims to study the molecules with potential antioxidant and iron-chelating properties for their capacity to inhibit the process of ferroptosis. Methods Our goal is to evaluate the effect of small molecules in modulating the ferroptosis pathway in cellular models. Test compounds are evaluated in different in vitro assays in PANC-1 cell lines sensitive to production of ferroptosis in presence of ferroptosis-inducing agent RSL3. The ability of test compounds to rescue cells from toxicity and oxidative stress upon treatment with RSL3 were evaluated. Various experiments like antiferroptotic cell viability study and DCFDA assays, , total GSH assay and study of different important biomarkers were carried out to judge the potential of compounds in modulating ferroptosis process and to provide antiferroptotic activity.
ResultsThe results show that dose dependent pretreatment of PANC-1 cells with multifunctional compounds D-512, D-583 and D-700 could protect the cells from toxicity of RSL-3 significantly compared to RSL-3 treated cells alone. The effect was found to be dose dependent. RSL-3 produces robust reactive oxygen species in PANC-1 cells. However, cells pre-treated with test compounds could dose dependently significantly reduce production of ROS upon exposure to RSL-3. The protection of level of GPX4 from exposure to RSL-3 by test compounds was evaluated. RSL3 reduces the level of GPX4. However, on pretreatment with the test compounds rescued the levels of GPX4. Similarly, SLC7A11 was rescued by pretreatment with the drug from toxic effects of RSL3. Another important biomarker for ferroptosis is NRF2. The loss of levels NRF2 upon treatment with RSL3 were restored by pretreatment with the test compounds. Finally, the levels of another marker FTH were unaltered after treatment with the test compounds. The test compound D-583 was able to protect the total GSH levels in the cells from toxic effect of RSL3 on treatment with our drugs. Conclusion In conclusion, the findings from this research present promising prospects for advancing innovative therapeutic approaches in the field of Parkinson\u27s disease and other neurodegenerative disorders. Small molecules dopamine agonist with iron chelation and antioxidant properties can modulate ferroptosis process and offer a potential avenue to disease modifying therapeutics in PD. Drug D-512, D-583 and D-700 showed promising antiferroptotic activity in PANC-1 cells against the RSL3 induced toxicity
Polychlorinated Biphenyl 126 Triggers Oxidative Stress And Reshapes The Gut Microbiome Across The Life-Span
No full textPolychlorinated biphenyl 126 (PCB 126), a dioxin-like pollutant, is known to cause oxidative stress, hepatotoxicity, and gut microbiota alterations. This study aimed to evaluate the effects of PCB 126 exposure on wild-type (WT) and flavin-containing monooxygenase 3 knockout (FMO3 KO) mice, focusing on liver pathology, oxidative stress, and gut microbiota changes. Additionally, we investigated how maternal PCB 126 exposure affects the gut microbiota of aged offspring. In the adult exposure model, male C57BL/6 WT and FMO3 KO mice were administered PCB 126 over 12 weeks. Hepatic histology, oxidative stress markers (F2-isoprostanes), and gut microbiota composition were analyzed. WT mice exhibited significant hepatotoxicity, characterized by macro- and microvesicular fat deposition in the liver, which was absent in FMO3 KO mice. WT mice also showed increased oxidative stress with elevated levels of 8-iso-15-keto PGE2, while FMO3 KO mice were not changed in response to PCB 126 exposure. PCB 126-induced changes in gut microbiota varied significantly between genotypes. In WT mice, Proteobacteria levels decreased following exposure, while they increased in FMO3 KO mice, indicating a genotype-specific microbiome response to environmental toxins. In the maternal exposure model, dams were exposed to PCB 126 during preconception, gestation, and lactation, with half of the groups also receiving exercise interventions. Offspring gut microbiota composition was assessed at 49 weeks of age using 16S rRNA sequencing. Maternal PCB 126 exposure significantly reduced microbial richness and diversity in offspring, regardless of diet or exercise. Specific taxa alterations were observed, including a depletion of Verrucomicrobiaceae and Akkermansia muciniphila, and an increase in Anaeroplasma. These microbial shifts suggest that maternal PCB 126 exposure may predispose offspring to chronic diseases later in life. Overall, the study highlights the differential response of FMO3 deletion in mitigating PCB 126-induced oxidative stress and liver damage, as well as the complex interactions between genetic factors and environmental toxins in shaping gut microbiota compared to WT mice. These findings underscore the importance of considering both genetic predispositions and early-life environmental exposures in assessing long-term health risks associated with PCB 126
Addressing Salon Segregation: Structural Racism In The Salon Industry
Get PDFIn the United States, hair salons remain unofficially segregated by race, reflecting broader patterns of systemic inequality with deep historical roots and contemporary manifestations. This dissertation examines Black hair not merely as a physical attribute but as a profound aspect of cultural identity and personal expression. It aims to illuminate the broader implications of salon segregation, challenging the notion that it is a natural outcome of human preferences or tendencies. Instead, it argues that this segregation is the result of entrenched social processes, deliberate structural racism, and discriminatory policies that cause significant harm to Black individuals. By presenting testimonials from Black individuals, this work gives voice to the lived experiences of those navigating segregated salon spaces. It underscores the importance of acknowledging and validating these experiences to combat contributory injustice and the epistemic violence that often silences marginalized voices. This dissertation also explores the historical evolution of Black hair care, from its African origins and treatment during slavery to its modern-day significance, demonstrating how historical biases continue to influence contemporary practices. Through a philosophical lens, the study examines the intersection of hairstyling and structural racism, argues that salon segregation results in material, social, and epistemic harms, and advocates for practical and philosophical solutions. The resolution of salon segregation touches on fundamental questions of dignity, identity, respect, and the right to belong—issues that resonate far beyond the salon chair