Mason Journals (George Mason Univ.)
Not a member yet
3256 research outputs found
Sort by
Atmospheric River frequency and strength in the Miocene over the North American west coast
No full textAtmospheric rivers (ARs) are essential to many communities, providing a seasonal source of rainfall, however, ARs are also hazardous through flooding and landslides. The double-sided nature of ARs mandates that we study them within the context of a warming climate. The Miocene Climatic Optimum (MCO; ~15Ma) provides a potential portrayal of a future warmer climate with CO2 levels ranging from 400-1000 ppm and temperatures 7-10°C warmer than today. To understand changes in AR characteristics we compare AR frequency and strength between Preindustrial and a suite of MCO simulations with varying CO2 concentration (400 and 560 ppm) using the Community Earth System Model (CESM1.3). To track ARs we used the Image-Processing based Atmospheric River Tracking (IPART) algorithms. Specific focus was given to ARs around the western coast of North America during November, as ARs are highly prevalent and make up a majority source of precipitation for the western North American winter. Over a ten-year average we find an increase of ~31% in the number of AR events from Preindustrial to MCO conditions. We also find a ~10.4% increase in integrated water vapor transport (IVT) strength, indicating significant potential increases in AR-associated precipitation. This study shows how ARs behave under much warmer climate systems and provides insights into future AR trends
Titanium-infused Nichrome Thin Films for Applications in Extreme Thermal Environments
No full textNichrome (NiCr) alloys are widely used in nanoelectronics and aerospace due to their relatively high electrical resistivity and ability to withstand high temperatures. Physical properties of NiCr, including high temperature stability, have been widely studied with sputtered NiCr thin films. However, literature lacks information regarding how adding foreign elements to NiCr (metallic impurities) improves or affects NiCr properties. This project explores the impact of Titanium (Ti), a conductive and corrosion-resistant metal, on physical properties of NiCr. Ti-infused NiCr alloys were synthesized using a co-sputtering technique while keeping the NiCr power at 250 W and varying the Ti power from 25 W to 150 W. Material characterization was performed by measuring the electrical resistivity, thickness, and composition by using a four-point probe, a stylus profilometer, and an energy-dispersive X-ray spectroscopy (EDS) technique, respectively. The study demonstrates that adding Ti to NiCr increases both the thickness and electrical resistivity, indicating a positive linear relationship while potentially improving the resistance of the film to heat corrosion. These unique properties of Ti-infused NiCr alloys may prove useful in both electrical and thermal engineering for extreme environments
Exploring the creative output of six LLMs measured by Torrance’s 4 components of creative thinking based on holistic and analytic cognitive styles
No full textAs Large Language Models (LLMs) become increasingly integrated into education, understanding their impact on creativity skills is essential. Existing research has explored LLMs in creative education, and how different cognitive styles judge and process AI-generated responses. Yet, research into the personalization of LLMs in responding to cognitive styles while solving creative tasks is minimal. This study explores how well LLMs support creative education, aiming to identify the most suitable LLM for creative output that can effectively adapt to the holistic-analytic dimension of cognitive styles. Using Torrance’s four components of creativity, three student researchers independently prompted six high-traffic general-use LLMs with a control, holistic, and analytic approach to generate multiple solutions to a specific problem. Solutions were scored with a standardized 1 to 5 rubric measuring relevance, reasonability, creativity, and adaption to cognitive styles (styling). Results show all LLMs achieved nearly perfect scores for relevance and reasonability (between 4.91 and 5). Within the four aspects of creativity, Gemini achieved the greatest mean score in elaboration (4.97), Grok in flexibility (4.27), ChatGPT in fluency (3.05), and Deepseek in originality (2.96). The overall creativity score shows Copilot (2.47) and Claude (2.62) performed worse compared to the other four LLMs. Based on the overall creativity score, Grok (3.28) showed the highest scoring performance, with statistical significance compared to all other LLMs (p < 0.05). In addition, Grok significantly outperformed all other LLMs in styling (p < 0.05) except Gemini (p = 0.064). No other LLMs showed statistical significance. Future work will expand the study with the use of additional specific problems prompted to these six LLMs to better determine one specific LLM for teaching creativity based on participants' cognitive style
Evaluating the Effectiveness of AI Assistants in Enhancing Conceptual Understanding of Machine Learning Classification Topics
No full textBridging the gap between knowledge and application is essential for effective learning, especially when it comes to machine-learning classification methods. While existing research has examined the effects of AI assistants on topics such as student performance and knowledge retention, limited work has been done in specifically addressing their impact on learning machine learning classification topics. Using a five-point difficulty scale, 20 peer researchers evaluated 15 fundamental classification concepts in a pre-experiment. Support Vector Machines (SVM), which are supervised machine learning algorithms that classify data by finding the optimal hyperplane that maximally separates different classes in a high-dimensional space, was rated as the most difficult topic (M = 3.38). This study focuses on analyzing the accuracy and variation between AI assistants in being used as a tool for understanding machine learning classification topics, specifically SVM. Using rubric-driven prompts for an SVM-based sentiment-shift detection task on the Amazon Fine Food Reviews dataset, we refined this discovery by comparing four AI assistants: ChatGPT, Google Gemini, Claude, and DeepSeek. Each assistant was evaluated on coding quality, reproducibility, explanatory clarity, and predictive performance (macro‑F1, accuracy, recall), with scores out of 100. ChatGPT achieved the highest macro‑F1 (0.82) and accuracy (0.84) and an overall rubric score of 88, significantly outperforming Google Gemini (F1 = 0.80, score = 79) and Claude (F1 = 0.59, score = 72). DeepSeek failed to show promising results, including the F1 score below 0.50. Furthermore, ChatGPT's answers showed excellent process organization and detailed clarity. These findings imply that the capacity of AI assistants to facilitate SVM learning differs significantly. Future research will investigate adaptive prompting techniques catered to learners' skill levels and expand this framework to additional classification methods
Internal Study of Artificial Intelligence Assistants (AIAs) to Identify the Most Effective AIA For Socratic Learning in Time Series Analysis
No full textSocratic-style learning is a form of self-regulated learning (SRL) that uses guided questioning and dialogue to promote deeper understanding. Although effective in traditional settings, facilitating Socratic-style learning in asynchronous environments remains a challenge, as there is a lack of tools capable of replicating Socratic dialogue. While artificial intelligence assistants (AIAs) are becoming increasingly prominent in education, most models are designed to provide direct answers rather than guiding the learner through Socratic dialogue. There is limited research on whether AIAs can support Socratic-style learning, particularly in complex subjects such as time-series analysis (TSA) where solving questions requires a deep conceptual understanding. This study investigates whether conversational AIAs can facilitate Socratic-style learning of TSA. Nine AIAs were tested with a Socratic prompt and ChatGPT-4o, Copilot (1.25063.108.0), and Gemini 2.5 Pro were identified as the only ones capable of following Socratic dialogue. Based on this pre-experiment, a rubric was created to evaluate the AIAs’ step-by-step reasoning, responsiveness, and Socratic engagement. Each identified AIA was given three realistic TSA problems (from undergraduate college courses) and assessed by 3 student researchers, with the scores being averaged across the three questions. From the experiment, Copilot, with an average score of 64.88, demonstrated limited instructional restraint, generated fewer exploratory questions, and loss of continuity with the initial prompt. Gemini 2.5 Pro achieved an average score of 85.19 and ChatGPT-4o achieved an average score of 81.00, demonstrating the strongest abilities to guide users through Socratic questioning. Since Gemini 2.5 Pro and ChatGPT-4o performed similarly, further experimentation may be conducted to identify whether Gemini 2.5 Pro or ChatGPT-4o is better suited for this task
Chain-of-thought and Tree-of-thought Analysis
No full textLarge Language Models (LLMs) like ChatGPT and Claude are increasingly being used to help student learning. However, the most effective prompting strategies for beginner programmers remain unknown. Two methods, Chain-of-Thought (CoT) and Tree-of-Thought (ToT), offer structured reasoning in different ways: CoT provides a step-by-step explanation, while ToT uses multiple reasoning paths. Although both of these methods have shown promise in helping improve AI performance, little to no research has examined their comparative impact on human learning. In this study, it was evaluated how CoT and ToT prompts influenced a beginner’s understanding of JavaScript sorting algorithms, including Bubble Sort, Selection Sort, Merge Sort, Insertion Sort, and Quick Sort. Using GPT-4o, selected because of its top benchmark performance in coding and reasoning, AI-driven lessons were generated and a five-category rubric—clarity, technical accuracy, inclusion of warnings, retention, and understanding—was used to assess their effectiveness. Three students with little to no prior experience in JavaScript participated in the experiment. While CoT scored a total of 66 points across the rubrics, ToT scored a total of 55 points. Participants reported that CoT responses were easier to follow, boosted confidence, and improved their ability to retain and apply the information learned. In contrast, while ToT did encourage broader thinking, it often introduced a cognitive load that hindered comprehension for beginners. These results indicate that CoT prompting clearly outperformed ToT prompting and is more effective for teaching algorithmic fundamentals to beginners. Limitations of this research include small sample size and limited content scope. In the future, a larger group of students could be gathered and taught using the two different reasoning mechanisms with tests and examinations determining the results. In addition to this, other programming topics other than sorting algorithms could also be explored for this research along with several other reasoning mechanisms
Multi-Camera 3D Behavioral Tracking of Rodents using OptiTrack and DANNCE Deep Learning Framework
No full textIn experimental neuroscience, understanding how neural activity correlates with behavior during spatial navigation tasks requires precise tracking of animal movement kinematics. This typically involves recording brain signals in animals, such as mice, as they explore mazes, alongside simultaneous video tracking to capture behavioral data. However, there is currently a lack of comprehensive pipelines that integrate camera calibration, labeling procedures, and 3D reconstruction into a seamless workflow.
In this project, we used a multi-camera system with four OptiTrack Flex 13 cameras arranged around a custom-designed experimental arena to capture high-resolution 3D rodent behavioral data. Camera calibration was performed using OptiTrack’s Motive software, employing wand-based motion calibration and ground-plane tools to obtain precise intrinsic and extrinsic parameters. These parameters were then applied in DANNCE, a markerless video-based 3D tracking system that integrates projective geometry and 3D convolutional neural networks (CNNs) to infer animal landmarks across camera views. Approximately 2–3 minutes of synchronized video frames were manually labeled in MATLAB to identify anatomical keypoints, which were triangulated into accurate 3D coordinates. These labeled frames formed the training dataset for the DANNCE deep learning framework, enabling it to predict detailed 3D skeletal representations from multi-view video for the remainder of the recordings.
Preliminary results demonstrated calibration errors consistently below 0.5 mm, indicating high spatial accuracy. The manual labeling and triangulation process successfully generated reliable 3D keypoints for model training. Initial DANNCE training showed promising accuracy in predicting rodent skeletal postures and movements from unlabeled video data, with further validation underway to enhance model performance and robustness. This integrated 3D tracking system offers a powerful tool for quantitative analysis of rodent behavior, with significant potential to advance neuroscience and behavioral research
Modeling of Squalene Epoxidase Mutations Suggest Subtle Structural Modifications Confer Resistance to Terbinafine
No full textFungal resistance in infections represents an escalating global health concern, particularly among elderly and immunocompromised populations. Terbinafine, a front-line fungicide targeting squalene epoxidase (SQLE) faces increasing resistance from strains featuring mutated SQLEs. Almost no structural information exists for wild-type (terbinafine sensitive) or mutant (terbinafine resistant) fungal SQLEs. While a human SQLE structure with a terbinafine analog is available, it features a significantly different sequence. Alphafold2 was used to generate models of wild-type and mutant SQLEs. These models were aligned with the human SQLE crystal structure using sequence and structural alignments to determine potential causes of terbinafine resistance. Structural analyses reveal that terbinafine resistance arises from point mutations that alter the space within the binding pocket causing either gaps that terbinafine can not fill or creating steric conflicts. Both cases result in improper interactions between terbinafine and the binding pocket that reduce binding affinity. The human SQLE also exhibits reduced affinity for terbinafine and can also be explained by similar principles. We find that even conserved residues exhibit different orientations which further influence resistance. Understanding these structural changes provides a foundation for developing new antifungal agents capable of targeting resistant SQLE variants
A Snail’s Tale: Assessment of Morphological Differences Between Introduced Freshwater Mystery Snails in North America
No full textThe introduction of non-indigenous species has led to a decline in biodiversity within freshwater ecosystems in North America. The species Cipangopaludina japonica and Cipangopaludina chinensis, also known as the Japanese and Chinese mystery snails (JMS and CMS), respectively, are invasive to North America. However, identification of the two species is complicated by their overlapping ranges and morphology, potentially requiring genetic testing. Using a morphology-based method of distinguishing between the two species would make identification more efficient and widely accessible than genetic testing. To assess morphological differences, we used Image J to measure genetically confirmed specimens (25 CMS, 41 JMS) collected from the United States and Canada, measuring a series of ratios between points on the snails’ shells. We found that CMS had a shell length (SL) to shell width (SW) ratio of 1.235 ±0.05, while JMS had a significantly larger SL/SW ratio of 1.358 ±0.07 (p <0.001). The ratio of SL to height from the major whorl downwards (bi) for CMS was 1.318 ±0.03, while JMS was significantly lower (1.357± 0.08) (p =0.003). Finally, the ratio of SL to whorl height (ae) for CMS was 1.948 ±0.06, significantly higher than JMS (1.863 ±0.14) (p = 0.004). A Principal Component Analysis showed that the SL/SW ratio was the most important when distinguishing between the two species. Although ratios differed significantly, overlap indicates that morphological methods of identification are ambiguous within a certain range (1.226 - 1.333 SL/SW, 1.272 - 1.376 SL/bi, 1.817 - 2.071 SL/ae), and genetic testing may still be required
Creating Distortion Frames in 2-D Animation via Motion Tracking and Trajectory Analysis
No full textTraditional animation necessitates the use of squash and stretch principles, where animators identify motion extremes and apply deformations to create believable character movement. However, the process is labor-intensive, requiring expertise and time investment. Current computer vision techniques allow for the creation of in-between frames in animation, but there is a gap in automated systems that can intelligently apply creative distortions such as squash and stretch based on motion analysis. This study presents an artificial-intelligence system that detects moving objects in animation sequences and applies contextually appropriate squash and stretch deformation. The methodology combines background subtraction using MOG2 for object detection, trajectory analysis for motion extreme identification, and physics-based deformation calculation. The system analyzes motion patterns to identify three key animation movements; peaks, impacts, and direction changes. Deformation parameters were calculated based on motion intensity, with squash factors ranging from 0.3 to 2.0. The processed animations showed enhanced visual appeal, properly applying squash and stretch effects, emphasizing dynamic movement while maintaining object volume. The system provides an enhancement to existing animator workflows, reducing the effort needed in planning squash or stretch while retaining high-quality animation