Mason Journals (George Mason Univ.)
Not a member yet
3256 research outputs found
Sort by
Analyzing the Relationship Between Project Success and Project Type for Open Source, Decentralized Finance and Web 3.0 Projects
No full textDecentralized applications and services are pivotal to the crypto industry. This research investigates the developmental history of over 600 open-source crypto projects using GitHub Archive data spanning from 2013 to 2023. Utilizing SQL scripts, the study extracted various project activities—such as watches, pull requests, pushes, commits, and branch creations—to construct a comprehensive history for each project. These projects were categorized into types, including crypto wallets, Dapps, L2 networks, tokens, and stablecoins, to identify trends in GitHub activity by project type. Additionally, projects were distinguished by their funding and governance models. The hypothesis posits that different project types may display distinct activity patterns, such as Dapps having more commits or token projects having more branches. Furthermore, this research examines the correlation between a project's governance and funding model and its success, measured by the level of user development and interaction on GitHub. It is hypothesized that more successful projects exhibit higher and sustained activity levels. This analysis aims to determine if certain funding and governance models, such as those of DAOs versus private companies, reliably correlate with project success, or if external factors play a more significant role. The findings provide insights into the governance of decentralized corporations and the impact of business models on project success, enhancing our understanding of the factors driving development and engagement in the crypto and open-source software communities.
This abstract is part of a collection in which the overarching large project under Dr. Jiasun Li was subdivided into discrete critical tasks that were carried out by multiple individuals or smaller teams. Abstracts in this collection read similarly given the shared project goals, but represent distinct tasks completed by the abstract authors towards finalizing the described analysis
Revolutionizing Inefficient Scheduling Systems by Adapting Offline Programming Techniques to Function in an Online Algorithm within a Python Application
No full textModern scheduling mechanisms fail to accommodate real-world time management constraints, often wasting time and preventing optimal usage. To solve this problem, two approaches may be used: Offline and Online Algorithms. Offline algorithms receive full input before creating output, while online algorithms receive input incrementally and make decisions without full data. While many studies cover offline algorithms, there is a lack of knowledge about online ones. We designed a program that analyzes key factors, including time slot popularity and reservation history, to make optimal choices. Users enter their desired time range for an activity, and the algorithm determines the best room. The efficiency of this room selection is measured by the greatest potential to meet the needs of future users. Preliminary data suggests that in comparison to both a random room selection as well as the default offline algorithm using linear programming (considered to be an optimal output as it is given the full data), the online algorithm developed performs around 250% better than a random one but only 75% as well as the offline one (based on number of users scheduled in total). With larger amounts of historical data, it is highly likely that this algorithm would be able to outperform current results, but in its current form, it still serves as a valuable scheduling tool and a foundation for future research
Analyzing of Funding and Governance Structures on GitHub Development Activities in Open Source Software Projects
No full textOpen-source software (OSS) projects are increasingly gaining prominence with the emergence of new funding and governance models. However, the relationship between these models and project development activity remains unclear. This study examines the possible connection between funding/governance systems and development metrics on GitHub, such as commit and watch counts, in OSS projects. The first part of the study examines data from approximately 600 OSS projects using BigQuery, encompassing repository names, dates, actor IDs, actor logins, and 28 different events, including total activities and distinct commits, from 2013 to the present and information on funding mechanisms (such as public token sales, crowdfunding, and product sales) and governance structures (like decentralized autonomous organizations (DAOs), centralized foundations, and private companies) alongside GitHub activity metrics. Data from GitHub Archive was retrieved using SQL in BigQuery to obtain project-level metrics, while project websites and GitHub repositories were reviewed to categorize funding and governance models and identify project types (such as cryptocurrency wallets, decentralized applications (DApps), and Layer 2 networks). By comparing GitHub metrics across different funding and governance categories, this study aims to identify patterns and trends in development activity. The findings will provide deeper insights into factors influencing OSS project success, such as the impact of financial resources from public token sales versus traditional product sales on developer activity. Additionally, this research will demonstrate how governance structures, such as community-driven DAOs versus centralized private companies, can affect development workflows and project longevity.
This abstract is part of a collection in which the overarching large project under Dr. Jiasun Li was subdivided into discrete critical tasks that were carried out by multiple individuals or smaller teams. Abstracts in this collection read similarly given the shared project goals, but represent distinct tasks completed by the abstract authors towards finalizing the described analysis
New nylon fiber protein extraction method advances hydroponic rhizosphere microbiome studies
No full textHydroponics, although an innovative and efficient method for growing crops, often provides conditions suitable for root rot, commonly caused by Pythium ultimum (syn. Globisporangium ultimum), a plant pathogen to which garnet red amaranth (Amaranthus tricolor) microgreens are particularly susceptible. This study explored the effects of two root rot biocontrol products on hydroponic A. tricolor microgreens. Six hydroponic tanks were set up with A. tricolor microgreen seeds on burlap and treated with Hydroguard (Bacillus amyloliquefaciens by Botanicare) and Orca (liquid mycorrhizae by Plant Success). The pH was maintained at an average of 6.1, the fertilizer concentration at 1.5 mS/cm3, and the temperature at 23 °C. Hydroponic water samples were collected from effluent, and proteins were then isolated using a newly developed nylon fiber technology dyed with Sudan IV red. Gel electrophoresis results indicated that the nylon protein extraction method successfully captured proteins from the water samples. Observation results showed that control tanks had greater root coverage on average compared to the treated roots. This suggests that the inoculants have a negative effect on root growth. However, these findings are confounded by the unintended inconsistencies in irrigation flow rate. More trials with reliable irrigation are needed to gauge the effects of different microbial inoculants on hydroponic microgreens. Nylon fiber extraction from effluent water presents a new easy method that can advance research in hydroponic pathogens and microbial treatments
Computational Analysis of Ion Velocities Along Coronal Loops
No full textAs the Sun generates heat within its hot, dense core, this energy radiates outwards towards the much cooler surface. As the surface is approached, the gas begins to convect. This movement of plasma generates a magnetic field around the Sun, which creates arching structures of hot plasma known as coronal loops above the Sun’s surface. These magnetic loops often reach much higher temperatures (over 10^6K) than the Sun’s surface despite being even further away from its core.
Coronal loops are a critical feature of the Sun’s active regions, where intense magnetic forces drive extreme solar weather events like solar flares and coronal mass ejections. Such events have the potential to cause geomagnetic storms on Earth and disrupt key technologies such as satellites, communications systems, and power grids. By studying coronal loops, we can better understand various aspects of the solar atmosphere, including extreme solar weather events that directly impact life on Earth.
To better understand the nature of coronal loops, our team analyzed solar data from the Hinode spacecraft’s Extreme-ultraviolet Imaging Spectrometer (EIS). We utilized a specialized Python library called EISPAC alongside existing public documentation. We successfully determined the ion velocities at various points along coronal loops at a variety of temperatures. These velocities can be used with computer models to better understand the physics of coronal loops. The results of our velocity measurements will be presented
Calculation and Comparison of Velocities of Magnetic Loops on Sun’s Surface Using Ions at Different Temperatures
No full textIt has been found that the surface of the Sun (the photosphere) has a lower temperature than the outer atmosphere (corona). As energy is generated in the 15-million-degree core due to nuclear fusion, it flows outwards to the photosphere and the corona. However, while the temperature of the photosphere is 5000 Kelvin, the temperature of the corona can be over 10^6 Kelvin. Currently, the heating mechanism that causes this drastic change in temperature is unknown. The answer to this seems to be in the magnetic loops that form below the photosphere and rise to the corona. The study of these magnetic loops on the surface can pave the way for a better understanding of the mechanisms that heat the outer atmosphere of the Sun. In addition, it can help provide insight into predicting solar flares and other solar activity. Specifically, the aim of our research is to produce velocity measurements that allow for a comparison with the results of computer models that employ different heating methods. Our research is based on EIS data from the Hinode spacecraft. Using Python and Python packages, we plotted points on the ultraviolet image of the Sun to identify a magnetic loop. These points were analyzed for different ions with different spectral lines. We fit each spectral line with a Gaussian function; from the resulting parameters, we calculated the velocities using the Doppler effect. Based on the calculated velocities, we presented velocity graphs as a function of ions of different temperatures
Application of Physics-Informed Neural Networks to Asthma Epidemiology
No full textAsthma is a chronic lung condition in which the airways become inflamed and narrow, and overproduce mucus, making breathing difficult. Asthma affects 262 million people worldwide, and as exposure to pollution is a key risk factor in developing conditions, increasing urbanization is often accompanied by an increase in asthma prevalence, particularly is lesser-developed regions. To examine the relationship between asthma and pollution, we constructed a system of ordinary differential equations using a compartmental model with susceptible, exposed, and infected components, as well as a pollutants component to act as a pathogen. The parameters of the model that describe the relationship between components are difficult to measure and are currently unavailable. Thus, a physics-informed neural network approach was taken to compute parameter values for a given dataset. The network was trained on artificially generated time series data of each component. The error on estimated parameters was calculated and optimized based on known input parameters to the system. Once the error is reduced via hyperparameter tuning, the network can compute accurate parameters given real-world time series data to allow for realistic modeling of asthma-pollution epidemiology
Improving infectious disease predictions through the use of metapopulation SIR modeling and graph convolutional neural networks
No full textGraph convolutional neural networks have shown tremendous promise in addressing data-intensive challenges in recent years. In particular, some attempts have been made to improve predictions of SIR models by incorporating human mobility between metapopulations and using graph approaches to estimate corresponding hyperparameters. In [1], researchers have found that a hybrid GCN-SIR approach outperformed existing methodologies when used on the data collected on a precinct level in Japan. In our work, we extend this approach to data collected from the continental US, adjusting for the differing mobility patterns and varying policy responses. Extensions and generalizations of the metapopulation GCN-SIR learning framework are proposed.
[1] Cao, Q., Jiang, R., Yang, C., Fan, Z., Song, X., Shibasaki, R., “MepoGNN: Metapopulation epidemic forecasting with graph neural networks’’, Joint European Conference on Machine Learning and Knowledge Discovery in Databases, pp. 453–468 (2022
Symptom-Based Prediction of Respiratory Diseases Risk in Relation to Air Pollution Exposure
No full textWith over half of the world's population exposed to rising air pollution levels, the risk of lung damage anddiseases like interstitial lung disease (ILD), pulmonary sarcoidosis, and chronic obstructive pulmonarydisease (COPD) increases. Early detection significantly enhances survival and recovery rates, yet researchusing air quality data for lung disease prediction in the U.S. is scarce. This study aims to develop apredictive model for lung disease incidence based on particulate matter (PM2.5) and ozone levels in theUnited States, promoting more frequent screenings and early detection.
We used daily PM2.5 and ozone measurements from the Environmental Protection Agency (EPA)alongside interstitial lung disease and COPD data for 2000, 2005, 2010, and 2014. Machine learningtechniques were employed for data preparation, analysis, model building, and training. Preprocessingsteps included merging and cleaning datasets to ensure data quality and robustness. The Spatial LagModel (SLM), Spatial Error Model (SEM), and Geographically Weighted Regression (GWR) were used toanalyze the spatial-temporal dependencies between PM2.5, ozone, and lung disease mortality.County-specific latitude and longitude measures were used to build the GWR.
Preliminary results confirmed a significant correlation between air pollutant levels and lung diseaseincidence. Our machine learning models effectively identified patterns and trends that can be leveragedfor early prediction of lung disease cases. The findings underscore the potential of utilizing air qualitydata for predicting lung disease incidence, facilitating early diagnosis and treatment. This researchprovides valuable insights into the health impacts of air pollution and highlights the necessity for furtherexploration in this area
Enhancing Digital Interactions with AI Agents for Greater Convenience and Efficiency
No full textAI agents are software entities capable of performing digital tasks autonomously, adapting to new challenges andenvironments without direct human control. Recently, AI agents powered by large language models like GPT-4 have beenintegrated into operating systems such as Ubuntu, Windows, and macOS. These agents can install apps, edit images,remove tracking devices from websites, and automate routine digital tasks. OSWorld is a platform that supports theresearch of these agents. In this project, we reproduced the OSWorld environment and analyzed the agents’ ability tocomplete digital tasks using GPT-4o. The research results showed that AI agents can be used in a wide range of digitalapplications, offering significant practical benefits and assistance to users in various settings