Digital Commons @ Texas A&M University-San Antonio
Not a member yet
3365 research outputs found
Sort by
Artificial Intelligence in Research: Compliance Implications
Get PDFWith new developments in Artificial Intelligence surfacing every day, Institutional Review Boards and Human Research Protection Programs have their work cut out for them. While the possibilities are endless and great, so too are the risks, and research administrators must learn to navigate this new terrain responsibly, ethically, and in full compliance with government and institutional standards
Evaluation of Schoenoplectus Californicus in an Artificial System for E. coli Removal
No full textGreen technologies and nature-based solutions like constructed wetlands have be come more popular due to their environmentally friendly approach to remediation, low cost of construction, and maintenance. Also, they have shown efficiency in removing a wide range of contaminants. However, choosing a suitable aquatic plant for this system is one of the questions that should be answered wisely. Schoenoplectus californicus– Giant Bulrush is an aquatic plant that was tested in this controlled constructed wetland, and the final concentration of the targeted microbial contamination, which is E. coli, was reduced remarkably. Also, for this study, soil was removed from the system, and instead, a hydroponic system was chosen to monitor the effectiveness of different variables on the system. According to the results, water quality was improved through the engineered constructed wetlands, especially in terms of the effectiveness of this aquatic plant in both removing E. coli and improving turbidity in the contaminated water, while other factors either remained the same or decreased throughout the five days of the experiment
A Short Proof of the Lemma of the Logarithmic Derivative in Several Complex Variables
No full textIn this work, we provide a concise proof of the logarithmic derivative lemma in Cn. We start by giving a complete proof of a critical result by Biancofiore and Stoll (Ann. of Math. Stud., No. 100, 29–45, 1981) via matrix theory, which is central to this area of study. Then, we follow Li (Trans. Amer. Math. Soc. 363, 6257–6267, 2011) and Ye (Math. Z. 222, 81–95, 1996) to present the shortest proof to date
Seeing Gray? Color Incongruity\u27s Effect on Facial Identification
No full textIdentification card (ID) screening occurs in everyday life prior to purchasing goods or gaining access to establishments. Unfortunately, contextual factors can bias face-matching (e.g., low prevalence effect, document bias, and size disparity). We focused specifically on color incongruity (i.e., comparing two identities is biased when one facial image is in color and the other is grayscale). Participants wore eye-tracking glasses while comparing 48 physical IDs (printed in color [congruent] or grayscale [incongruent]) to a color facial video. Participants answered the question, “Are these pictures of the same person?” using a scale from 1 Definitely No to 6 Definitely Yes. Data collection is ongoing. However, we predict differences in patterns of response bias and eye movements for incongruent color compared to congruent color trials. This preventable incongruency has implications of fraud and criminality that can affect security, law enforcement, and private citizens. Influencing the future of ID verification systems and protocols
Cardiac Responses to Klebsiella pneumoniae Infection: Investigating its Role in Sepsis and Endocarditis
No full textSepsis causes nearly 11 million deaths across the globe and about 350,000 per year in the United States. Sepsis can be induced by a number of viral, fungal, and bacterial pathogens. Klebsiella pneumoniae is particularly adept at disseminating through the blood to many distal sites from the lung. Klebsiella pneumoniae is a Gram-negative bacterium that can cause pneumonia, meningitis, endocarditis, wound infections, and is a leading cause of bacterial sepsis. The purpose of these studies is to observe how Klebsiella pneumoniae affects the heart. These studies will be important in elucidating how cardiomyocytes respond to infection with Klebsiella pneumoniae. Two approaches will be taken to uncover how these cells respond to infection. Heart tissue samples will be harvested from uninfected and infected mice to observe the in vivo response. Briefly, mice will be infected and tissues will be harvested at different times post infection. The students will assist in processing the tissues and perform immunofluorescence assays to help delineate immune responses in the heart at different times post infection. The second approach involves directly infecting immortalized rodent cardiac cell lines to study the susceptibility of these cells in vitro. Our goal is to define how resident heart cells respond to Klebsiella induced endocarditis
MEASURING AND IMPROVING THE EFFICIENCY OF PYTHON CODE GENERATED BY LLMS USING COT PROMPTING AND FINE-TUNING
No full textWith the advanced AI technologies, the role of Large Language Models (LLMs) has grown rapidly for software development with generating the code that is functionally correct, solving complex problems, and debugging existing code. However, LLMs often produce inefficient code with unnecessary logic, hallucinated content, and errors. This research measures the efficiency of Python code generated by GPT-4o-Mini, GPT-3.5-Turbo, and GPT-4-Turbo models using execution time, memory usage, and maximum memory usage while maintaining correctness. Using EffiBench datasets on Google’s Vertex AI Workbench with different machine configurations, the study uses the seed parameter for consistency and optimization techniques like Chain-of-Thought (CoT) prompting and fine-tuning. Except for GPT-4-Turbo, the results show that CoT prompting improves efficiency metrics for GPT-4o-Mini and GPT-3.5-Turbo. GPT-4o-Mini was selected for fine-tuning due to its better results with CoT prompt and its costeffectiveness but fine-tuning compromises accuracy and efficiency. Overall, high-CPU machine configurations, along with GPT-4o-Mini and CoT prompting, improves the efficiency and correctness of LLM generated Python code in resource-intensive scenarios
Optimizing Executive Cash Bonuses: The Nonlinear Impact of Executive Cash Bonuses and Strategic Fit
No full textThis study examines how aligning executive cash bonuses with firm strategic orientation influences firm performance. While prior research mainly focuses on stock-based incentives and linear effects, we address critical gaps by analyzing cash bonuses and exploring nonlinear relationships using polynomial regression and response surface analysis.We adopt Miles and Snow’s typology and focus on prospector and defender firms, representing the two ends of the strategic continuum. Our findings indicate that aligning executive cash bonuses and firm strategy significantly enhances future performance. Neither cash nor stock incentives alone, without strategic alignment, substantially improve performance. Polynomial regression results reveal an inverted U-shaped relationship between executive cash bonuses and firm performance. Performance improves with cash bonuses up to roughly 50% of total compensation, beyond which further increases become counterproductive, potentially demotivating executives or encouraging inefficient strategies, especially in efficiency-oriented defender firms. Moreover, the surface analysis shows that the effect of the optimal bonus threshold differs by strategic type, with innovation-oriented prospector firms reacting more sensitively to cash bonus levels than defenders. These findings advance compensation and management control literature by highlighting the importance of strategic alignment and emphasizing a nonlinear relationship between executive cash bonuses and firm performance. Practically, our research offers important insights into designing effective compensation structures tailored to a firm’s strategic orientation
The Impacts of Ethnic Studies
No full textThis mixed-methods study examines the impact of Mexican American Studies (MAS) courses on high school students in South Texas, focusing on academic success and personal development. While research in California and Arizona has demonstrated the positive effects of ethnic studies on GPA, attendance, and graduation rates, Texas has yet to implement comprehensive studies assessing the efficacy of such courses. This study seeks to fill that gap by employing a transformative mixed-methods design, integrating quantitative data—such as standardized test scores, GPA, and attendance records—with qualitative insights from student and educator interviews. Grounded in Critical Race Theory, LatCrit, and Chicana Feminist epistemology, this research investigates how MAS influences students\u27 academic performance, sense of identity, and engagement in school. The findings aim to inform policymakers and educators about the necessity of ethnic studies in Texas K-12 education, advocating for culturally relevant and sustaining pedagogies that address systemic inequities and support student success
Preliminary investigation of changes in fish morphology as related to turbidity
No full textHuman activities increase turbidity in freshwater environments. Shoaling, a behavior with numerous benefits for fishes, is known to be hindered by turbidity. We investigated whether turbidity also induces morphological changes in fishes, specifically alterations in eye size or body shape. Using ImageJ software, we analyzed images of 218 Red Shiners (RS, Cyprinella lutrensis) and 135 Western Mosquitofish (WM, Gambusia affinis) provided by the San Antonio River Authority. Regression or correlation analyses revealed that relative eye and pupil size in RS significantly increased with turbidity (Nephelometric Turbidity Units), a trend not observed in WM. Similarly, RS exhibited a significantly deeper relative body depth with increased turbidity, while WM did not. These findings suggest that turbidity\u27s impact extends beyond behavioral changes, influencing fish morphology. This likely reflects the loss of shoaling benefits and the necessity for fish to adapt to this environmental stressor
Leveraging AI for the Identification of Aquatic Invertebrate Species
No full textThe identification of similar-looking aquatic invertebrate species is an especially difficult task due to their subtle morphological differences. To address this challenge, we developed a Roboflow model trained on a dataset that includes images from iNaturalist as well as images from our own collection. The model is designed to identify 15 distinct species of aquatic invertebrates native to the greater San Antonio region. Here, we evaluate the model’s performance, in terms of its accuracy in detecting and correctly labeling each of these species. Additionally, we are working towards deploying this model in a mobile format such that it is accessible for field use, allowing for the identification of species in real-time. Ultimately, this tool benefits both researchers and students by enabling the accurate, efficient, and rapid identification of aquatic invertebrates