Digital Commons @ the Georgia Academy of Science
Not a member yet
1290 research outputs found
Sort by
PERCOLATION THEORY: ANALYTICAL SOLUTIONS AND NUMERICAL SOLUTIONS USING MONTE-CARLO AND AI SEARCH METHODS (IN PROGRESS)
No full textPercolation theory is one of the simplest models that can accurately describe phase transitions in complex physical systems. Examples of phase transitions range from the macroscopic, such as forest fire propagation, water through porous mediums, and oil dispersion, to microscopic phenomena that include quantum phase transitions and magnetic transitions. While the one-dimensional lattice can be solved explicitly for percolation threshold, mean cluster size, and correlation length, etc., various lattice structures in higher dimensions do not yield explicit solutions due to intense scaling properties which require the use of numerical approximations. Historically, methods such as the Hoshen-Kopelman algorithm were utilized for numerical solutions, but through the devlopment of C++ code that utilizes Monte-Carlo methods, dynamic data structures, and artificial intelligence search methods, best case time and memory complexity is greatly reduced. This presentation will discuss introductory concepts of percolation theory, the development of simulations, and how obtained results are analogous to important physical properties in two and three-dimensional systems
DESIGN OF A HIGH VACUUM CHAMBER AND MICROWAVE FABRY-PEROT CAVITY USING COMPUTER AIDED DESIGN SOFTWARE **
No full textThe proposed microwave spectrometer design, utilizing a Balle-Flygare type cavity Fourier transform, will be modeled on SolidWorks to provide a comprehensive blueprint in the manufacturing process. The Fabry-Perot cavity resonator, integrated with torispherical cap mirrors, will be digitally fabricated, optimizing the cost-effective vacuum chamber closure. Various torispherical domes, such as DIN 28011, ASME standards, and high crowns (80/10, 80/6, 90/8), will be assessed for reflector diameter, volume, and focus length. Different kinds of ASME standard flanges will be added to the chamber, and have the potential to be used in several applications, such as the use of diffusion pumps. This integrated approach ensures a more complex, yet precise representation of the design in the transition from concept to physical construction
BAND STOP BEHAVIOR OF METAMATERIAL
No full textAccording to Smith et.al [1990], a metamaterial composed of thin infinitely long metal rods arranged in a host medium with a periodic fashion exhibits “high pass” behavior. Therefore, it can be predicted from circuit point of view that if infinitely long metal rods are replaced with very short thin metal rods, the resulting medium will exhibit “low pass” behavior. However, if this replacement is done with finite sized metal rods or strips, the resulting medium behavior will be of stop band type because of series inductance (L) – Capacitance (C) resonance phenomenon. In this research, 3D electromagnetic simulation software HFSS was used to model a host dielectric material with thin finite sized metal strips on it to prove the band stop behavior. An experiment was also conducted to demonstrate this band stop or band suppression property. This work was supported by Georgia Space Grant Consortium
MODIFIED MICROSTRIP LINE BASED NONINVASIVE BIOSENSORS IN MICROWAVE SPECTRUM**
No full textNoninvasive, high-sensitivity and low-cost biosensors are an emerging area of interest for industrial and medical applications. Such sensors operate based on the material\u27s response to the applied electromagnetic field. Microstrip line-based transmission lines can generate significantly strong electromagnetic field in its vicinity. We propose a modified microstrip structure-based microwave transmission line capable of sensing fluid presence by applying electromagnetic field
EFFECT OF AMENDMENTS ON THE TRANSPORT AND LEACHING OF LEAD IN POLLUTED SOILS **
No full textMining is essential for acquiring a lot of the natural resources that humans require in their daily lives. Unfortunately, mining activities have resulted in widespread contamination of soils and negatively impacted people and the environment around the world. This project is part of a larger NSF IRES project and assesses the impacts of various soil amendments on the bioavailability and mobility of lead in soils. The objective of the study is to understand how these amendments affect the leaching of heavy metals and nutrients from polluted soils into groundwater. A greenhouse pot-study was conducted using heavy metal-contaminated soils, biochar, compost, and hyperaccumulator plants to investigate the influence of soil amendments on heavy metal leaching. X-ray Fluorescence Spectrometry was used to analyze the initial and final concentrations of lead in the soils, while spectrophotometry and chromatography were employed to analyze the leachate collected. Then a study was conducted in the field to investigate naturally contaminated sites in Zambia. Intact soil cores for laboratory column experiments were collected from the contaminated sites. Leachate from the soil cores was collected and analyzed for lead and nutrients. Preliminary results suggest that compost retains lead in the soils by about 30% when compared to unamended soils and other treatments. Being able to determine which amendments minimize the leaching of lead can be used to develop best management practices and guidelines for amending mining wastelands and tailing dumps. This material is based upon work supported by the National Science Foundation under Grant No. 2107177
COMPUTER-MEDIATED VERSUS FACE-TO-FACE DYNAMICS OF SELF-DISCLOSURE**
No full textThe current study examines how self-disclosure functions in computer-mediated communication (CMC) versus face-to-face (FtF) communication. Various research theories, including online disinhibition, self-awareness effects, and hyperpersonal communication, have indicated that inherent differences in CMC environments lead to increased frequency and intimacy of self-disclosure online. However, contradictory evidence in the literature prompts a need for further analysis of these findings. In the current study, we indirectly manipulated levels of online disinhibition and self-awareness by placing participants in various CMC conditions designed to elicit differing levels of these variables and compared their resultant levels of self-disclosure to those of participants in a control FtF condition. In this experimental design, we had dyads discuss a dilemma in one of three CMC condition groups: text-based chat (high disinhibition, low self-awareness), voice calling (moderate disinhibition, moderate self-awareness), and voice calling with webcam (low disinhibition, high self-awareness) or a control FtF (very low disinhibition, very high self-awareness) condition. A team of coders measured self-disclosure frequency through video and textual analysis. Questionnaires were administered to measure online disinhibition and self-awareness effects. At submission, the study was in progress and approximately 10% of the data had been acquired. Additional data collection continues and an ANOVA will be used for data analysis
USING COGNITIVE PSYCHOLOGY TO PROBE AI SOCIAL BIAS IN GPT-4**
No full textHow do Artificial Intelligence (AI) large language models structure the human world? And how does that compare to what humans do? Cognitive psychologists have long studied how humans represent the social world in our mind--for example, by relying on social category labels, such as race and gender, to make quick predictions about individual personal traits. Such mental shortcuts reflect inaccurate perceptions of the world and often lead to social prejudice and discrimination. While human rationality is bounded by limited mental resources, highly capable large language models, such as OpenAI’s Generative Pre-trained Transformer 4 (GPT-4), can process vast amounts of information quickly and efficiently, showcasing a level of cognitive capacity that can potentially surpass the constraints of human rationality. This project aims to apply the cognitive framework of psychological essentialism to investigate whether GPT-4 exhibits social essentialist bias similar to humans, as a way to explore the underpinning of AI social bias and identify areas where large language models mirror or rise above human irrationality. We will use a meta-prompt to instruct GPT-4 to evaluate a range of social categories along 6 essentialism items on a 9-point Likert Scale, with higher ratings indicating stronger essentialist bias. Through OpenAI\u27s API implementation, we will generate 150 responses in GPT-4 (sample size comparable to previous data collected with human subjects) in the Python language. We will calculate mean essentialist ratings and compare them with the scale mid-point (5.0) using one-sample t-test to examine whether GPT-4 displays an overall essentialist bias. We will also conduct generalized linear mixed-effects models to examine whether GPT-4\u27s responses differentiate from human responses and vary as a function of social domains (i.e., race, gender, nationality, etc.). Potential findings from this project will inform the responsible development and deployment of AI technologies in human-AI collaboration, particularly in decision-making processes typically prone to social biases
THE EFFECTS OF AIR QUALITY ON SOIL COMPOSITION IN GEORGIA, U.S.A.**
No full textAir and soil are two major components of a healthy ecosystem for plants and animals, which makes understanding and protecting these resources important for land management officials. In our research, we aimed to understand the relationship between air quality and soil composition in Georgia, U.S.A., by looking for correlations between the two among urban and rural locations where air pollutants may differ. We hypothesized that there is a correlation between air quality and soil composition. To examine potential correlations, in the fall of 2022 we collected soil samples within 50 m of eleven Environmental Protection Agency (EPA) air monitoring stations located from Northwest to Central Georgia. EPA air quality data from 2011-2022, showed average particulate matter concentrations differed across our eleven field sites, ranging from 8.25 ppm in rural areas to 10.25 ppm in urban areas. We used x-ray fluorescence analysis (XRF) to determine soil composition and will look for correlations between soil composition and air quality. The results will provide information about how air quality affects soil variance near Atlanta’s urban center to the rural areas of the state. Because soil and air play a major role in the water supply to cities and agricultural production, understanding the relationship between air quality and soil composition can help us mitigate future risk of unwanted pollutant contamination
RECRUITMENT AND ENROLLMENT OF PREGNANT WOMEN WITH SUBSTANCE ABUSE AND THE NEWBORN **
No full textThe HEALthy Brain and Child Development study (HBCD) is an ongoing prospective longitudinal study of approximately 7500 pregnant mothers and their newborn babies across 25 sites in the United States. The study digs deep into the assessments of the brain, evaluating the cognitive and emotional development of children from birth through the first 10 years of childhood. Twenty-five percent of the sample includes children who have a history of prenatal exposure to substances of abuse (alcohol, cigarettes, opiates, and marijuana). For a longitudinal study, research participant recruitment and retention are incredibly challenging, despite the enormous commitment of the researchers. This study aimed to deepen the knowledge on recruitment and retention of pregnant women in birth cohort studies and expand the insight into a more effective research engagement. The study evaluated the relationship between participant recruitment and the amount of effort needed to recruit the participants after they showed some primary interest at one enrollment site. The amount of the researchers’ engagement efforts were recorded on a weekly basis through unsolicited cold emails, phone calls, and texts The amount of patient recruitments were also recorded. The time interval between the first outreach from the researcher to a potential participant and the day when the participant completed the consent form was also recorded. Current data analysis suggests that researchers average 2.87 contacts (via emails, texts or calls) before a participant signs up. Also, it takes 10.25 days on average for a potential participant to enroll in the study. This one-and-a-half-week delay in enrollment aligns well with the weekly engagement efforts by the researchers, providing a reliable and consistent prediction for the overall pattern of study enrollment. Up to 75% of the participants were enrolled within 2 weeks, and a threshold of 29.125 days indicates that further contacts are no longer necessary
THE IMPACT OF LOW-BODY WEIGHT ON ESTIMATING AGE-AT-DEATH AND SENESCENCE OF FEATURES ON THE AURICULAR SURFACE OF THE PELVIC SACROILIAC ARTICULATION**
No full textForensic anthropologists are commonly tasked with constructing a biological profile when attempting to identify an individual based on skeletal remains. During this process, age estimation is commonly accomplished by scoring the auricular surface of the os coxae. However, this process has encountered interobserver error due to biological variation in height, weight and occupational stress. It has been previously determined that increased body mass index, or BMI, levels can result the overestimation of individual ages from the scoring of the auricular surface compared to their age at the time of death. However, less is known about the impact of low BMI on estimating age-at-death from the auricular surface. In a sample of 150 individuals from the Bass Skeletal Collection at the University of Tennessee-Knoxville, three age groups (30-49, 50-64, and 65-80) were examined and their sacroiliac joints scored using the Buckberry & Chamberlain system. Using mean ages from the composite scoring system used in Buckberry & Chamberlain, each of the three groups were designated a mean age and standard deviation using a 95% Confidence Interval. The 30-49 age group of known ages possesses a mean estimated age of 36.89; and a standard deviation of 1.10. The 50-64 age group of known ages exhibits a mean estimated age of 48.50 and a standard deviation of 0.89. Finally, the 65-80 age group of known ages produces a mean estimated age of 58.39 and a standard deviation of 0.92. This research indicates individuals with low BMI are often underestimated with respect to age based on the scoring of the auricular surface. The implication is that body weight, when available, should be taken into account when estimating age-at-death using the auricular surface to improve the accuracy and completeness of the biological profile