Clayton State Digital Repository
Not a member yet
1706 research outputs found
Sort by
Construction progress of Detector of Unusual Cosmic-ray casKades
Full text linkHigh Energy Physics (HEP) is a field that has still has many mysteries that need to be solved. An open question is about the origin and composition of the Ultra-high Energy Cosmic Rays (UHECRs). These cosmic rays originate well outside our planet, may even be outside of the galaxy. They are messengers that could help us better understand the universe around us and provide insight into the fundamental building blocks of our universe. The primary goal of the Detector of Unusual Cosmic casKades, is to detect and verify the existence of unusual cosmic events. Moreover, it can help innovate EAS (Extensive Atmospheric Shower) analysis methods. This poster aims to highlight developments of the detector system, instrument calibrations and other activities conducted at Clayton State University
Prototype Setup for the DUCK
Full text linkThis article covers the prototyping activities for the DUCK (Detector system of Unusual Cosmic ray casKades). The primary goal for the DUCK system is to observe the unusual cosmic events reported by other collaborations, and to look at the possibilities of adding innovations to the EAS (Extensive Atmospheric Shower) analysis methods of the EAS disk measurements at the observation level. The prototyping process has helped to choose between various design possibilities in the process of the optimization of the individual detector design
In Search of "Old Pete": A Critical Reassessment of Gen. James Longstreet's Constructed Childhood
Full text linkThis thesis examines historiographical inaccuracies in the portrayal of General James Longstreet’s childhood and highlights how these errors have shaped misconceptions about his adult character. Through a critical analysis of primary and secondary sources, it identifies twelve key inaccuracies introduced by historians and reassesses the formative influences on Longstreet’s childhood development which emphasizes his unwavering sense of duty as a central driving force. Contrary to the prevailing narrative that credits his uncle Augustus B. Longstreet and states' rights ideology with shaping his persona, this study argues that Longstreet was primarily influenced by his parents, siblings, and his close relationship with his enslaved childhood nurse, Daniel. Primary sources, including letters and personal recollections, reveal how these familial ties instilled in him a profound sense of responsibility that became evident in his military career and post-war actions. His valor in the Mexican War and pivotal role in the Civil War illustrate how this commitment to duty informed his decisions and transcended the political ideologies of his time. Moreover, his advocacy for reconciliation during Reconstruction reflects a vision of national unity rooted in duty and loyalty rather than regional interests. This work calls for a reevaluation of Longstreet's childhood and its impact on his adult nature and urges future research to explore the complexities of his early influences and their overall significance in shaping his legacy in American history.M.A
Implementing Newton’s method in python for root-finding in differentiable functions
Full text linkNewton’s Method is an iterative technique that uses tangent lines to approximate roots of real-valued, differentiable functions. This paper explores the method’s application as a strategy for finding real roots when exact solutions are difficult to obtain algebraically. The method was implemented in Python and applied to a variety of functions, including a low-degree polynomial, a high-degree polynomial, a trigonometric function, and an exponential function. The approximation process starts by evaluating each function at integer values between –10 and 10. If a function evaluated to zero at any of these values, a root is found immediately. Otherwise, the Intermediate Value Theorem is used to locate intervals where the function changed sign, indicating the presence of a root. One of these values is then selected as an initial approximation for Newton’s Method. Using the iterative formula, each initial guess is refined to converge toward the actual root. The method produced accurate approximations across all function types when an initial guess was sufficiently close to the true root, demonstrating the efficiency of Newton’s Method in iterative root-finding
Preliminaries of a Photon-based Hardware Random Number Generator Design and Data Analysis Methods
Full text linkToday, when computer-based devices take over our lives, security becomes of the utmost concern. HRNG, or hardware random number generators, are extensively used in the digital world for security purposes as well as in the science world as a source of high-quality randomness for the models and simulations. Existing HRNG are either extremely costly, or slow and of questionable quality of data. This publication describes a simple design of the HRNG based on the low-number photon absorption by a detector (a photo-multiplier tube of a silicon-based photodetector) that can provide a large volume of high-quality random numbers. The different options of processing and the testing of quality of the generator output are described
The Star System : A Novel Method of Encoding and Indexing Chinese Logographs
No full textThe topic of this study is Chinese character encoding, and the coded number sets produced are designated as Star numbers. This research is focused on creating a new system of identifying and indexing Chinese characters by numerically encoding them to produce a new bilingual dictionary. Chinese logographs can be represented by sets of two-dimensional line-segments encompassing the strokes and sub-strokes that compose the character. The compositional linear feature orientations of line-segments that constitute a logograph can be used to extract a five-digit number set that represents the character. The sum of line-segment orientations in each character can be categorized, producing a 5-digit numerical code that then becomes a direct numerical representation of the character. In the experiments produced by this research, the proposed collation of Chinese logographs was divided into two parts. Based on a standard set of coding rules, Star numbers were extracted from each character in a database of glyphs by visually counting and categorizing each line-segment orientation in the logograph. The various Star numbers were then lexically collated to examine clustering effects, assuming that too much clustering of unique characters around individual Star numbers would render the system impractical in a bilingual dictionary setting. The most extensive database created and tested, 6500 characters of Level 1 and 2 of The Table of General Standard Chinese Characters, produced the following results: 1) over 37% of the Star numbers tested contained a single character cluster, 2) almost 80% of the Star number codes included five or fewer character clusters, 3) over 95% of the Star numbers contained ten or fewer glyphs per unique Star number, 4) almost 99% of the Star codes included fifteen or fewer glyphs per cluster, and 5) the largest group of characters clustered around a single Star number consisted of twenty-three logographs. The results of this study suggest that the Star system approach of encoding and indexing Chinese characters may be an effective method for lexically collating the glyphs into a bilingual dictionary or another referencing system.M.A
Skin Cancer Awareness and Detection
Full text linkThis thesis explores the application of big data and machine learning technologies in the early detection of skin cancer, particularly melanoma, through digital health data. Traditional methods have struggled to handle the large, rapidly changing datasets associated with health monitoring, making advanced computational techniques increasingly essential. The research focuses on the development of a machine learning-based software, currently under development at the CS/IT Department of Clayton State University as part of a funded NSF project. This software, integrated into mobile applications for Android and Mac platforms, allows for secure image uploads and provides predictions of skin cancer likelihood, aiding in early diagnosis. The thesis aims to collect, analyze, and tabulate skin cancer detection data across various population segments to refine the software and enhance its diagnostic accuracy and potential for personalized prevention.The rapid increase in digital health data has led to new areas of research in healthcare and data sciences. Traditional methods of handling health data have struggled because they can't manage the huge, fast-moving, and diverse amounts of data that are constantly changing. Skin cancer, particularly melanoma, represents a significant public health challenge due to factors such as increased ultraviolet (UV) radiation exposure and evolving lifestyle patterns. The application of big data and machine learning technologies offers promising advancements in the early detection of skin cancer by processing and analyzing extensive datasets, which include patient histories, environmental exposures, and genetic predispositions. Machine learning algorithms, particularly those focused on dermatological image recognition, enable the identification of skin lesions with high precision, thus facilitating the timely diagnosis of melanoma. Furthermore, predictive analytics models can identify individuals at heightened risk, potentially enabling early interventions and more personalized preventive strategies. The integration of big data and advanced computational techniques into skin cancer detection holds the potential to significantly enhance early diagnosis, treatment outcomes, and overall prevention efforts. The software using the latest machine learning techniques is under development at the CS/IT Department at Clayton State University as a funded NSF project. The purpose of this thesis is to collect, tabulate, and analyze skin cancer detection data among various segments of population. In addition to the initial testing and detection is carried out by using the software smartphone application software are developed for Android and Mac platforms. The apps allow for secure and private upload of images predicting various types of skin cancer with a percentage likelihood.M.S
Book of Abstracts
Full text linkThis publication is the collection of all abstracts of works presented at the 4th Annual College of STEM symposium
Efficient Subgraph Search in Time-Dependent Networks
Full text linkSubgraph search is a problem that has long been studied in computer science and a number of algorithms have been designed to preform this task completely. Variations on this problem have included multi-graphs, colored edges and vertices, and hypergraphs among others. A network can be described as a graph, and as that network changes over time, these changes can also be modeled as a graph. In this paper we show how existing algorithms can be made more efficient when preforming subgraph search on a time dependent network.M.S