University of Illinois at Chicago
University of Illinois at Chicago: UIC INDIGO (INtellectual property in DIGital form available online in an Open environment)Not a member yet
21439 research outputs found
Sort by
Pseudo-Orthogonal Arrays: Definition, Construction, Comparison, and Application in Designs of Experiments
No full textInspired by orthogonal arrays OA(n,k,s,t), arrays whose inner product between any two column vectors being 0 are discussed. Such arrays can be divided into two categories: One can be derived from orthogonal arrays OA(n,k,s,t) with strength t≥2; the other have random experiment points spreading in the experimental space, thus are more flexible but may not be balanced regarding the level combinations, and we denote them by pseudo-orthogonal arrays (pOA).
Orthogonal Latin hypercubes can be treated as a class of pOAs. Introduced by Ye (1998), it take advantage of both Latin hypercube designs and property of orthogonality. Similar to the idea of OA, we can also generalize the Orthogonal Latin hypercube to fixed-level and mixed-level pOAs. With a more flexible structure, pseudo-orthogonal arrays have relatively higher factor-to-run ratios. The existence conditions and construction methods for fixed-level and mixed-level pseudo-orthogonal arrays are discussed and proposed. The relationship between the proposed criteria for construction algorithm and other optimality criteria is discussed.
Further, when a pseudo-orthogonal array cannot be found, an array with weak correlations between different factors are constructed to approximate the pseudo-orthogonal array. How and why the correlations should be controlled are discussed. We also give some examples of the application of pseudo-orthogonal arrays and their approximation in designs of experiments
Autism from Siblings' Perceptions: "Don't Be Thankful You Don't Have a Sibling With a Disability."
No full textThe purpose of this qualitative interview study was to explore the perspectives and experiences of neurotypical (NT) Arab American (AA) school-aged children who have a sibling with autism spectrum disorder (ASD). Given the growing number of Arab Americans in the United States and the limited literature on NT siblings of children with ASD, there is a rising need to investigate this unique cultural group and develop culturally responsive practices to meet their needs. This study offers valuable insights for researchers, educators, service providers, practitioners, and families of children with ASD. Eleven NT AA school-aged children aged 8 to 12 participated in semi-structured interviews to share their experiences with their siblings. Participants described their sibling relationships, needs, strategies for promoting social, academic, and emotional well-being, and support systems for themselves and their siblings in school and the community. Data analysis through grounded theory revealed that sibling dynamics among NT AA children with siblings who have ASD share key similarities with those of NT siblings from diverse cultural backgrounds. Six themes emerged from the data: (1) descriptions of the sibling relationships, (2) unique and complex bonds between siblings, (3) roles of participant siblings, (4) diverse needs of participants and their siblings, (5) sources of strength, and (6) support strategies utilized by participants. Participants highlighted unique perspectives shaped by their roles, yet common relational dynamics persisted, emphasizing shared characteristics with NT siblings across cultures. Additionally, family, faith, and peer support emerged as critical factors contributing to the well-being of these groups. This study underscores the critical need to enhance support services for NT AA siblings of children with ASD, as well as for other NT siblings from marginalized and culturally diverse backgrounds
Synthesis and Evaluation of Antiaddictive Natural Products
No full textThis dissertation seeks to provide data in support of four main claims: (i) molecular targets exist that, by inducing a variety of physiological changes, bring about the subjective and reinforcing effects of drugs of abuse; (ii) when these targets are acted upon by ligands, the subjective and reinforcing effects of drugs are sometimes mitigated; (iii) such ligands sometimes come from natural sources; and (iv) these natural products can be synthesized, derivatized, and evaluated in the lab. The novel synthesis and derivatization of the Aristotelia scaffold generated a library of analogs which deepened our understanding of its SAR at the α3β4 nAChR. The extraction and semi-synthetic derivatization of voacangine enabled the access of several iboga alkaloids which were evaluated for their ability to bind to DAT and SERT. Together, the synthesis and evaluate of these antiaddictive natural products contributed to basic science research that may lay the foundation for the development of a first-in-class drug to treat PSUD
GDPR Analysis and Automated Compliance Checking
No full textData is everywhere and is produced daily in huge quantities.
Not surprisingly, data spreads at the same pace of the links between data and individuals. Companies are more and more interested in collecting as much personal data as possible and serious concerns about data privacy arise.
All over the world, governments are facing the issue of protecting citizens from those who collect and process their personal data; in recent years, a number of privacy regulations have been enacted in Europe and US. The enacting of privacy laws has triggered the need for compliance processes, that still rely on manual efforts and therefore are costly, resource intensive and may lack coverage. The demand for automated tools is highly felt and spreaded among privacy champions.
In this work we'll focus on the General Data Protection Regulation (GDPR), a unique, general, comprehensive guideline for each and every privacy related matter.
First of all, we have carried out a dependency analysis on GDPR clauses that contain references to other articles or paragraphs.
The purpose of the analysis is to check, using a computer scientist rather than a jurist approach, whether cyclic dependencies occur.
On the way to compliance monitoring automation, we translate the legal text of privacy clauses into a machine-readable language.
We have exploited the results of the dependency analysis to improve a user-friendly language, that seeks a compromise between the extraordinary expressiveness of the natural language and the limitations imposed by a formal language.
Then, we enforce the translated privacy policies through a specific software framework.
The core component of such framework is a static analyzer, that checks the compliance of data analysis programs with formal privacy specifications.
We modified the static analyzer, to accept the formal specs resulting from the translation, and we adapted it to the specific needs of a Data Protection Officer.
We have set ourselves the ambitious goal of creating a realistic scenario: formalizing the significant clauses of the GDPR and using the analyzer to check the compliance of real open-source applications.
We passed satisfactory tests, both in terms of functionality and in terms of readability
Investigating the Role of Metabolite Signaling in Primary Metastasis of High Grade Serous Ovarian Cancer
No full textHigh grade serous ovarian cancer (HGSOC) is the most lethal and most common histotype of ovarian cancer. Unfortunately, many patients present in late stages of disease due to a lack of screening modalities. This lack of early monitoring is partly due to a deficit of understanding in the field about the early pathologic processes in HGSOC. Recently the field has come to understand that HGSOC can originate in fallopian tube epithelium, but how transformed cells migrate to and colonize the ovary is yet unknown. Hence, the bulk of this dissertation is focused on identifying signals that are involved in the primary metastasis of tumorigenic fallopian tube epithelial cells to the ovary.
We utilized imaging mass spectrometry to identify that tumorigenic fallopian tube epithelial cells and human ovarian cancer cells secrete a protein called SPARC which enables the ovary to release norepinephrine. We have also determined that norepinephrine can increase invasion in tumorigenic fallopian tube epithelial cells and increase epithelial to mesenchymal transition hallmark proteins in human ovarian cancer cells. Further, we have identified that different mutations in fallopian tube epithelial cells cause the ovary to release distinct sets of metabolites as measured by LC-MS/MS. Additionally, we have identified that kaempferol, a flavonoid found in nature, has partial progestogenic activity in mice and may represent a potential non-synthetic treatment for symptoms for which progesterone is a common treatment.
Ultimately, this body of work attempts to understand and address the current lack of clarity surrounding primary metastasis in HGSOC in hopes of identifying future avenues of prevention or treatment
<b>A Methodology for Identifying High Priority Areas in Need of Green Space: A Case Study in Chicago</b>
No full textThe benefits of green space (GS) on environmental and public health have been well documented for decades by researchers and urban planners. Historically, Chicago and other metropolitan areas have long prioritized the creation and maintenance of these spaces. However, studies show that these GS benefits are disproportionally allocated with major deficiencies in socially vulnerable communities. This finding underlines the need to establish a GS allocation methodology that will identify high priority areas in need of GS augmentation based on a two dimensional approach. This will simultaneously account for the lack of GS and the socioeconomic characteristics of the communities in need of GS augmentation. The establishment of such a methodology as an equitable GS allocation decision support tool is fully justified by the extreme weather events and the pollution burden impacting disproportionally socially vulnerable communities.The proposed two-dimensional approach utilizes a coincidence matrix that combines two key dimensions: green space (GS) coverage across the area and socioeconomic metrics that highlight inequalities among census tracts (CTs). This method is exemplified using the City of Chicago. The GS coverage is represented at the CT level, derived from publicly available datasets, while the socioeconomic dimension is informed by the Racial and Ethnic Minority Status component of the CDC’s social vulnerability index (SVI). Through this matrix, Chicago’s CTs are classified based on their GS coverage and SVI ranking, resulting in the identification of 378 high-priority CTs that require greening interventions. Further analysis of these priority areas using the coincidence matrix identified 154 CTs categorized as critical, where both metrics fall into the lowest GS coverage and highest SVI ranking levels.This paper has been recognized to be the best among those presented by young professionals at the 118th A&WMA Annual Conference & Exhibition.</p
Ensuring Relevance to Practice: Examining Implementation of an Influenza Vaccine Toolkit in Texas' Public Health Region 11
No full textOne hundred and fourteen Texas Vaccine for Children (TVFC) healthcare providers in Public Health Region 11 (PHR 11), Texas were sent the Texas Department of State Health Services (DSHS) influenza toolkit by mail in May 2022. The rationale for the creation of the DSHS influenza toolkit was to provide an educational tool for healthcare providers that could bring attention to and encourage vaccines to increase influenza vaccination rates. Toolkits offer practical advice on issues to facilitate wide-spread adoption of concepts or best practices across a profession (American Library Association, n.d.) and toolkit development and distribution should include literature reviews that uncover relevant, current, evidence-based practice (EBP) that could be presented to healthcare providers in a helpful and useful way (Cifuentes, 2022; Nowalk et al., 2012). The adaptable documents of a toolkit can increase the use of evidence-based interventions and inform and aid in implementation (Thoele et al., 2020).When developing a toolkit, gathering input from the users of the toolkit is a very important activity to provide a practical and accessible toolkit, as well as identifying and eliminating barriers for the users to promote self-efficacy. There was no logic model produced for the DSHS toolkit project, so this study created a theoretical logic model that proposes elements, actions and outcomes involved in the implementation of a vaccine promotional toolkit for providers, and more specifically, an influenza toolkit. The logic model indicates that resource requirements may be multilevel, including internal to DSHS, inter-organizational and community-based.A case study using exploratory, sequential, mixed methods design in a developmental evaluation (DE) framework was used for this study. This quantitative and qualitative methods approach was used to answer the study aims of how the toolkit was used and if portions of the toolkit were adapted or modified to meet their needs.</p
Fashioning Futures, Designing Dreams: Gender, Class and Dress in Kathmandu
No full textAre fashionable dressing practices reflective of class belonging, or can they be cleverly leveraged to create lasting class mobility for individuals with the knowledge and ability to know how to properly deploy them? This dissertation attempts to answer this question by examining the role of modern material culture in the lives of lower-middle-class women who have chosen to study fashion design at one of the most affordable of Kathmandu’s now numerous “fashion institutes.” My focus is on how these unmarried women (in their late teens to early twenties) use dress—as object and symbol—to mediate and contest the pressures they experience between expectations to be traditional (and marriageable within their class and caste communities) and modern, between the necessity to be frugal and the desire to claim a place in the city’s fashioned and fashionable middle-class public sphere. The project’s epicenter is the Namuna College of Fashion Technology in a middle-class Kathmandu suburb, where I conducted nine months of extensive autoethnographic research alongside women enrolled in Namuna’s fashion degree course. The research consisted of participant observation while students completed their coursework, extended to many aspects of informants' lives, and incorporated formal interviews and media analysis. The dissertation chapters detail the modern-day dressing practices of women in Kathmandu, examining how the supply and distribution of clothing in the city affects young women's choices about what to wear. The chapters also consider the way that kinship, gendered moral expectations, and social surveillance inform the way that young women present themselves and the way they agentively use this self-presentation to meet their own future goals. Most importantly, the dissertation highlights the way that young women use fashion as a practice and fashion education as a form of cultural capital to make claims about what—and ultimately, who—is fashionable in the city to influence the dressing practices of others. Simultaneously, they are also attempting to create a space for themselves in higher class households where their special talents for self-presentation and fashion consumption can become evidence of their family’s material success and global modernity
Evaluating Large Language Models for Turboshaft Engine Torque Prediction
No full textRecent advancements in deep learning (DL) have introduced transformative opportunities for time series forecasting, particularly through the use of transformer-based architectures. Among these, Large Language Models (LLMs), originally developed for natural language processing (NLP) tasks, have demonstrated strong capabilities in modeling sequences, learning from limited data, and integrating heterogeneous inputs. Their scalability and generalization abilities suggest that they could be leveraged beyond textual applications, offering new perspectives in industrial forecasting tasks.
This study investigates the potential of LLMs for time series forecasting in the aerospace sector, with a focus on turboshaft engine torque prediction. Accurate prediction of engine torque is vital for maintaining the reliability, efficiency, and safety of helicopter operations. Although statistical models and DL architectures have achieved notable success in this domain, the application of LLMs remains relatively unexplored.
To address this gap, this research evaluated and compared the performance of four different models: GPT-2 and ChatGPT (general-purpose LLMs), TimeGPT (a specialized transformer-based model for time series data), and a conventional time series transformer model. The models are assessed on a real-world dataset derived from a Bell 407 helicopter’s Health and Usage Monitoring System (HUMS), incorporating multiple exogenous variables relevant to engine performance.
The evaluation includes forecasting accuracy, robustness to reduced training data (few-shot learning), and classification performance with respect to operational torque thresholds. Our findings indicate that GPT-2 and TimeGPT achieve a strong predictive accuracy. ChatGPT demonstrates potential as a prompt-based LLM.
This work offers new insights into the applicability of LLMs to time series tasks in aerospace. It also highlights the importance of prompt design, data modality adaptation, and architectural specialization for achieving competitive performance. Overall, LLMs represent a promising direction for developing scalable and adaptable forecasting solutions in industrial domains
Generative Models Driven Graph Outlier Detection
No full textGraph data are pervasive across various domains, including social networks, biological networks, and communication systems. The detection of outliers in graph data—substructures that significantly deviate from the norm—is crucial for uncovering fraudulent activities, network vulnerabilities, and novel patterns. However, traditional outlier detection techniques often fall short of effectively modeling the complex non-Euclidean graph data. Recently, generative models have exhibited extraordinary performance on image and language data, but their capabilities in graph outlier detection remain largely underexplored. In this dissertation, I systematically investigate the capabilities of generative models in the context of graph outlier detection, including four published works and one ongoing work. Specifically, first two works introduce a comprehensive graph outlier detection library and a benchmark on existing graph outlier detection algorithms. The third work presents a diffusion model–based data augmentation for addressing class imbalance in graph outlier detection. The fourth work explores scalable global spatiotemporal attention in graph Transformers for graph outlier detection. The last work focuses on the fake news detection capabilities of large language models and constructs a large-scale real-world text-attributed graph dataset for graph outlier detection
Data sources:
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇