Volgenau School of Engineering Graduate Research

Permanent URI for this collection

https://hdl.handle.net/1920/11421

Browse

Now showing 1 - 20 of 72

Employing UAF Inter-Domain Traceability for Performance and Effectiveness Evaluation
(2023-03-25) Alghamdi, Ahmad; Torkjazi, Mohammadreza; Davila-Andino, Arturo J.; Zaidi, Abbas K.
We propose a step-by-step Model-Based Systems Engineering (MBSE) process for the creation and simulation of an executable Unified Architecture Framework (UAF) model for evaluation purposes. The roll-up of Technical Performance Measures (TPMs) to Measures of Effectiveness (MOEs) is necessary for such a process, and has not been documented for the UAF. This paper is the first attempt to address this gap by demonstrating how interdependencies between these technical measures can be traced across the domains of a UAF architecture according to the ISO/IEC/IEEE 15288:2015 standard, the guidelines from the INCOSE Systems Engineering Handbook, and the UAF Enterprise Architecture Guide. The proposed process employs traceability and parametric diagrams within the UAF to produce an executable model that aids in evaluating the effectiveness of a system’s architecture. Additionally, we describe how to build a simulation within the UAF to assess a parametric diagram containing random values of TPMs. The process identifies UAF views, their constituent model elements, and the relationships that are required to build this model. We also present an illustrative example of a forest firefighting system to demonstrate the implementation and effectiveness of the proposed process. This paper is intended as a resource for systems engineering practitioners.
Creating a Digital Twin of an Insider Threat Detection Enterprise Using Model-Based Systems Engineering
(2022 IEEE International Systems Conference (SysCon), 2022-04) Lee, James; Alghamdi, Ahmad; Zaidi, Abbas K.
Inference Enterprise Modeling (IEM) is a methodology developed to address test and evaluation limitations that insider threat detection enterprises face due to a lack of ground truth and/or missing data. IEM uses a collection of statistical, data processing, analysis, and machine learning techniques to estimate and forecast the performance of these enterprises. As part of developing the IEM method, models satisfying various detection system evaluation requirements were created. In this work, we extend IEM as a digital twin generation technique by representing modeled processes as executable UML Activity Diagrams and tracing solution processes to problem requirements using ontologies. Using the proposed framework, we can rapidly prototype a digital twin of a detection system that can also be imported and executed in systems engineering simulation software tools such as Cameo Enterprise Architecture Simulation Toolkit. Cyber security and threat detection is a continuous process that requires regular maintenance and testing throughout its lifecycle, but there often exists access issues for sensitive and private data and proprietary detection model details to perform adequate test and evaluation activities in the live production environment. To solve this issue, organizations can use a digital twin technique to create a real-time virtual counterpart of the physical system. We describe a method for creating digital twins of live and/or hypothetical insider threat detection enterprises for the purpose of performing test and evaluation activities on continuous monitoring systems that are sensitive to disruptions. In this work, we use UML Activity Diagrams to leverage the integrated simulation capabilities of Model-Based Systems Engineering (MBSE).
Creating a Digital Twin of an Insider Threat Detection Enterprise using Model Based Systems Engineering
(AI4SE and SE4AI Workshop, 2021-10-20) Lee, James; Alghamdi, Ahmad; Zaidi, Abbas
UAF Strategic Planning for Enterprises
(IEEE, 2022-11-24) Torkjazi, Mohammadreza; Davila-Andino, Arturo J.; Alghamdi, Ahmad; Zaidi, Abbas K.
In this paper, we propose an extension to the Unified Architecture Framework (UAF) Profile to leverage the potential of the widely used SWOT analysis in modeling the UAF strategic viewpoint of an enterprise or System of Systems (SoS). This novel methodology assists with capturing the capabilities an enterprise needs to achieve its goals through the SWOT analysis. Capabilities are key elements in the UAF Strategic viewpoint. Previous works for identifying capabilities only showcase the resulting strategic views without interpreting their ties to the UAF strategic elements. The absence of a common methodology for incorporating strategic planning into the UAF is a potential deterrent for enterprises to accept the UAF. To address this gap, we propose using the SWOT analysis as an established method of strategic planning that has been used for many years in various enterprises. The four SWOT elements, i.e., Strength, Weakness, Opportunity, and Threat, are comparable to the UAF strategic elements Challenge and Opportunity which are the basis for capturing the enterprise capabilities. We extend the UAF profile to enable SWOT elements in the UAF and provide dependency relationships to illustrate traceability between SWOT elements and capabilities. This paper provides two key contributions. First, this work extends the UAF Profile to identify capabilities from SWOT analysis which takes both internal and external conditions of the enterprise into account. Second, we develop the methodology to create the necessary views which trace the SWOT elements to the stakeholders and capabilities. The proposed method is one that enterprises and SoS managers can employ to adopt UAF with minimal to no disruption to their current business processes. We also demonstrate a seamless integration of the SWOT analysis into the UAF. Finally, we use an illustrative example of a hypothetical enterprise to demonstrate the new SWOT diagrams enabled through the proposed extension.
Data Analysis for Fraud Detection in Finance
(2022-06-04) Almutairi, Raghad; Godavarthi, Abhishek; Reddy Kotha, Arthi
Credit card use is not always the best way to use for payments, but the most demonstrable payment mode is through the credit card for both offline as well as for online payments, which can result in deficit of funds. As the online shopping is booming it helps in rendering the cashless payment modes. It can be used at shopping's, paying rent, paying utilities bill, internet bill, travel and transportation, entertainment, food. Using for all these things there is a chance of fraud transactions for a credit card, hence there is more risk. There are many types of fraudulent detections most of the banks and institutions are preferring fraud detection applications.it has become very hard to find out the fraud detections, After the transaction is done there is a chance of detecting fraudulent transactions in the manual business processing system. In real time the bunco transactions are done with real transactions, but it seems not to be sufficient for detecting . Machine learning and data science both are playing a very important role in identifying the fraud detections. This study uses data science and machine learning for detecting the fraud detection to demonstrate various modellings. The problem enables the transactions of the previously done transaction data.
Machine Learning Application in Health
(2022-06-10) Alshabana, Ghadah; Tran, Thao; Chitimalla, Ashritha; Thompson, Michael
Coronavirus can be transmitted through the air by close proximity to infected persons. Commercial aircraft are a likely way to both transmit the virus among passengers and move the virus between locations. The importance of learning about where and how coronavirus has entered the United States will help further our understanding of the disease. Air travelers can come from countries or areas with a high rate of infection and may very well be at risk of being exposed to the virus. Therefore, as they reach the United States, the virus could easily spread. On our analysis, we utilized machine learning to determine if the number of flights into the Washington DC Metro Area had an effect on the number of cases and deaths reported in the city and surrounding area.
ML models for Customer Relationship Analysis in Finance
(2022-06-08) Jia, Nan; Bagam, Lahari; Fabijanczyk, Patricia
The purpose of this research project was to analyze customer complaint data from financial institutions and identify areas of opportunity for these institutions to improve their customer satisfaction rate. In addition to pointing out areas for improvement, this paper also looks into similar research and tries to understand if themes found in this analysis are consistent with those done by other researchers. Banking is an essential piece to everyday life for all people across the world. Banks need to ensure that their products and processes are simple and accessible to all. Although banks have a monopoly on our financial needs their desire to retain existing customers and gain new ones drives the necessity of providing excellent and timely customer service. The study was conducted using a dataset of over two million customer complaint records and examining what were the top three financial institutions receiving complaints and which products received them. In addition, other aspects of complaints such as state of origination was also looked at. Analysis was done using machine learning, python, tableau and other tools to show the data points and their correlation. Understanding the top financial institution's methods of handling customer complaints, we are able to make recommendations for further product improvements to increase customer satisfaction. Concluding the research project is a list of challenges and opportunities for further research projects. In addition, there are recommendations for the financial institutions investigated in this project on how to move forward from analyzing customer complaint data.
Using Machine Learning Algorithms to Improve Road Safety
(2022-06-08) Kotikalapudi, Lakshmi Praneetha; Chao, Yen Ling; Reddy, Sri Siddhartha
Transportation facilities are becoming more developed as society develops, and people's travel demand is increasing, but so are the traffic safety issues that arise as a result. And car accidents are a major issue all over the world. The cost of traffic fatalities and driver injuries has a significant impact on society. The use of machine learning techniques in the field of traffic accidents is becoming increasingly popular. Machine learning classifiers are used instead of traditional data mining techniques to produce better results and accuracy. As a result, this project conducts research on existing work related to accident prediction using machine learning. We will use crash data and weather data to train machine learning models to predict crash severity and reduce crashes.
Data Analysis to Analyze Mental Health in Global Pandemic
(2022-06-08) Antonova, Helen; Liao, Spencer; Chandra Sai Pamidimukkala, Eswara
According to the 2021 Report from the World Health Organization (WHO), more than 700,000 people have taken their life. Suicide can be prevented but so far most of the efforts to do so have fallen short. However, the use of machine learning and artificial intelligence offers new opportunities to increase the accuracy level of prediction and aid the goal of suicide prevention. This paper reviews literature concerning the machine learning methods used to help identify various risk factors and help prevent suicide. This paper also presents our research and analysis findings which were used to identify various suicide risk factors and additional analysis of whether there are any correlations or variations in the risk factors from pre and post-pandemic datasets regarding suicide rates. This is especially important during times of high stress, such as a worldwide pandemic and quarantine. The dataset(s) obtained from WHO suggest that high levels of risk factor identification are possible, and this paper and the analysis serve as supporting research and guide to aid in the continued ambitious goal of suicide prevention worldwide
Neurological Manifestation of COVID-19: An Updated Literature Review
(2022-05) Eltayeb, Sohaib; Peixoto, Nathalia
SARS-CoV-2, which causes the disease known as Coronavirus Disease 2019 (COVID-19), is a novel coronavirus that arose in Wuhan, China in 2019. Within a short time, it rapidly spread across the world and has been declared a global pandemic by the World Health Organization (WHO) due to its severe morbidity and mortality rate. This virus left many scientists and biomedical engineers perplexed due to the various uncertainties about its infection rate as new COVID-19 variants arise. However, attempts to contain the virus are ongoing all over the world. Some of the most common symptoms of COVID- 19 include fever, dry cough, and fatigue. However, some physicians in affected areas have discovered that some patients that were diagnosed with COVID-19 did not exhibit these expected respiratory symptoms at the time of diagnosis, but rather these patients displayed only neurological symptoms as their initial symptoms. For instance, the symptoms range from non-specific to more particular, such as headaches or dizziness which were one of the more common symptoms, to more complicated symptom onset such as convulsions, cerebrovascular and peripheral diseases.
EEG- EMG-Based Interface For Upper Limb Exoskeleton – A Review
(2022-05) De Marzi, Laura; Peixoto, Nathalia
The second most common cause of death in the world is cerebrovascular accident or stroke, and rehabilitation plays an important role to help the survivors of such accidents. Rehabilitation exercises are essential to speed up the process of recovery and regain independence, not only for post stroke cases but, also, for every patient who suffers of other neuromuscular diseases, such as spinal cord injuries or multiple sclerosis. The aging of the population, the increase of accident, and therefore, the increase of quality and quantity of rehabilitation needed, have led to the development of new techniques and assistance methods for recovery. Exoskeleton robotic devices have been developed to help the rehabilitation process, complementing the manual work of therapists. What is needed for an efficient and smooth implementation of this device is an advance interface between the wearable robot and the human. In this paper we have presented and analyzed two possible control input signals for exoskeletons, specifically electromyography (EMG) and electroencephalography (EEG). We’ve delved deeper into these two techniques, studying their advantages and disadvantages. Advantages are for example their inherent intuitiveness and effectiveness. On the other hand there is high inter-subject variability of the EMG, and the non-invasiveness and high temporal resolution but relatively poor spatial resolution of the EEG technique. The purpose of this review is to study and contrast the two main techniques when used as brain machine interface for the control of exoskeletons.
Combinatorial Trends of Tissue Engineering for Peripheral Nerve Regeneration
(2022-05) Dando, Emma
Peripheral nerve damage is frequently seen due to injury or illnesses, like diabetes. Despite its prevalence and the fact that many patients with less serious injuries have good clinical outcomes, many patients do not fully recover sensation and in many cases, use of the affected area. For this reason, there has been extensive research into improving or replacing the current treatment options. Many tissue engineering solutions focus on peripheral nerve injury or damage caused specifically by external trauma. The aim of this paper is to list and summarize the primary areas of research for tissue engineering approaches to peripheral nerve regeneration. Moreover, the focus is on the increasing awareness that no single tissue engineering technique is currently capable of providing optimal healing and regeneration for peripheral nerve damage and may never be fully capable of providing complete regeneration. Instead, clinical outcomes may be improved by combining these techniques in multifaceted approaches some of which include combining growth factors and nerve guidance conduits.
Early Detection of Epilepsy Seizures using Machine Learning Algorithms
(2022-05) Rios, Cristian (Chris)
According to the World Health Organization (WHO), one of the most common neurological diseases is epilepsy with over 50 million people globally diagnosed to date. Epilepsy is defined as a chronic noncommunicable disease of the brain that affects the central nervous system and can result in sudden seizures. The early detection of seizures would allow suffering patients an opportunity to better manage their seizures and improve overall quality of life. In this paper, we aim to detect seizures using different machine learning (ML) algorithms using non-invasive EEG data of epilepsy patients through a publicly available database. We evaluated different signal pre-processing methods to allow for extraction of optimal features that include descriptive statistical parameters, frequency-based trend changes, power spectrum density, and entropy. For some classifiers, optimal features were passed through a Principal Component Analysis (PCA) algorithm to minimize the number of features used and to reduce the dimensionality of the ML model. ML models evaluated for seizure detection included Logistic Regression, Decision Tree, SVM, KNN, Naïve Bayes, and Neural Networks. A confusion matrix was generated for optimal models to compare accuracy, precision, and specificity. Results demonstrated the models were promising but the top three models with highest accuracy were DT, SVM, and NN. Given these results, ML models such as SVM or Neural Networks, can be utilized to create seizure detecting systems that are portable and use EEG data to detect the onset of seizures. Even though the highest accuracy observed was approximately 82.5%, the value of machine learning models show merit for solving seizure detection challenges. In future research studies, we plan to evaluate additional open databases to enhance these ML models further and increase accuracy to a minimum level of 95% to be evaluated in real-world applications.
Effect of Psychedelic (LSD and Psilocybin) Use in Treating Mental Disorders like Major Depressive Disorder (MDD)
(2022-05) Kahn, Zehra
Major Depressive Disorder (MDD) is a widespread problem throughout the United States with no shortage of treatments. Psilocybin, a serotonin agonist, is treatment for MDD that showed great potential but wasn’t explored much until recently. Psilocybin, when administered in micro doses along with therapy, showed great curative potential for MDD, ameliorating symptoms for long periods of time. Psilocybin has also shown to work quickly and was not addictive to the patients. This review looks at studies testing psilocybin as a treatment for MDD.
A Study of Epileptic Seizure Detection using Machine Learning Algorithms
(2022-05) Kamaraju, Rajeev; Peixoto, Nathalia
This paper focuses on studying epileptic seizure detection using machine learning algorithms. Algorithms like Naïve Bayes, Logistic Regression, Stochastic Gradient Descent, KNearest Neighbour, Decision trees and random forests have been studied. For each of the classifier, many performance metrics have been computed and Area Under Curve (AUC) has been chosen as our performance metric. The paper also introduces the possibility of detecting epileptic seizures using Neural networks.
Modeling the Effects of Dehydration on Cellular Growth and Wound Repair
(2021-06-15) Dando, Emma
Dehydration is a common problem among athletes and the elderly, both of whom get injured more frequently than the average adult. Dehydration has been linked to reduced circulation, a reduced amount of water being absorbed into the cellular matrix, increased inflammation, decreased blood pressure, decreased volume of plasma, increased heart rate, decreased cellular density, and increase adhesion constant, depending on the severity and duration of dehydration. A comprehensive and functioning model describing the impacts of dehydration on tissue growth could be used when studying and designing treatment plans for groups vulnerable to dehydration and to inform decisions on environments for growing cell cultures and tissue in vitro. Previous models have described hydration and water replacement in the human body or the process of tissue growth or wound repair. This model integrates the effects of dehydration with the parameters of wound repair to create a more comprehensive model. This resulting model is a combination and modification of a continuum and a partial differential equation model. The continuum model is modified to consider more of the effects of dehydration on the tissue with the intent to increase the accuracy of the model. The overlapping parameters and relationships are used to link the models together. The combined model is then modified with unitless parameters that represent the severity and duration of dehydration to create the final model. The model found that while the simulated subject is fully hydrated, intralayer elastic couplings are the largest factor that increases cell density, being to a power of nine. The influence of the tangential diffusion of cells is only to a power of one and comparatively uninfluential. The grouped effect of cell crowding and cell synthesis and apoptosis to decrease cellular density is to a power of negative nine, so mathematically it balances out the effect of the elastic couplings. The influence of the tangential velocity of cells is the next most influential component to decrease the cellular density, at a power of five. During severe dehydration, the cell crowding component and the elastic component changed by a power of four, the cell crowding component becoming less negative and the elastic couplings component increasing. Both the tangential diffusion and tangential velocity decreased by a power of two. Unexpectedly, the cellular density changed very little with varying simulations of dehydration. Despite this, the model shows that dehydration slows the rate of wound healing and it suggests that the severity of dehydration is more detrimental than the duration of dehydration.
Bert Model for Social Media Bot Detection
(2022-03) Heidari, Maryam; Jones, James H Jr.
Millions of online posts about different topics and products are shared on popular social media platforms. One use of this content is to provide crowd-sourced information about a specific topic, event, or product. However, this use raises an important question: what percentage of the information available through these services is trustworthy? In particular, might some of this information be generated by a machine, i.e., a ``bot" instead of a human? Bots can be, and often are, purposely designed to generate enough volume to skew an apparent trend or position on a topic, yet the consumer of such content cannot easily distinguish a bot post from a human post. This paper introduces a new model that uses Bidirectional Encoder Representations from Transformers (Google Bert) for sentiment classification of tweets to identify topic-independent features for the social media bot detection model. Using a Natural Language Processing approach to derive topic-independent features for the new bot detection model distinguishes this work from previous bot detection models. We achieve 94\% accuracy classifying the contents of data set Cresci \cite{cresci-etal-2017-paradigm}as generated by a bot or a human, where the most accurate prior work achieved an accuracy of 92\%.
Online User Profiling to Detect Social Bots on Twitter
(2021-03) Heidari, Maryam; Jones, James H. R.; Uzuner, Ozlem
Social media platforms can expose influential trends in many aspects of everyday life. However, the trends they represent can be contaminated by disinformation. Social bots are one of the significant sources of disinformation in social media. Social bots can pose serious cyber threats to society and public opinion. This research aims to develop machine learning models to detect bots based on the extracted user's profile from a Tweet's text. Online users' profile shows the user's personal information, such as age, gender, education, and personality. In this work, the user's profile is constructed based on the user's online posts. This work's main contribution is three-fold: First, we aim to improve bot detection through machine learning models based on the user's personal information generated by the user's online comments. The similarity of personal information when comparing two online posts makes it difficult to differentiate a bot from a human user. However, in this research, we turn personal information similarity among two online posts as an advantage for the new bot detection model. The new proposed model for bot detection creates user profiles based on personal information such as age, personality, gender, education from user's online posts, and introduces a machine learning model to detect social bots with high prediction accuracy based on personal information. Second, create a new public data set that shows the user's profile for more than 6900 Twitter accounts in the Cresci 2017\cite{cresci-etal-2017-paradigm} data set. All user's profiles are extracted from the online user's posts on Twitter. Third, for the first time, this paper uses a deep contextualized word embedding model, ELMO\cite{peters-2018-deep}, for social media bot detection task.
Machine learning models for Prediction of the need for future Covid-19 vaccine booster
(2021-04) Marzook, Ahmad Al; Xu, Ge; Jagannath, Prajna Shetty
About 4 million Virginia citizens are fully vaccinated against COVID-19. According to medical data from nations like Israel and the United Kingdom, all those people may require another shot in the near future, at least when keeping transmis- sible diseases at bay. Health officials are looking at whether a booster shot is needed to ensure the vaccine’s effectiveness. To prepare it for predictions, we will develop a new simple logistic regression model. Make prediction models using classification and probability. With the vaccine immune response lifetime taken into account, we will run our model to forecast the completely vaccinated number timeline and observe when it achieves the herd immunity percentage of the overall Virginia population. The percentage of people that need to be vaccinated to attain COVID- 19 herd immunity is still up for dispute. The goal of this study was to forecast when everyone would be fully immunized. COVID-19 vaccination loses nearly half of its defense antibodies every 108 days, according to Elie Dolgin. As a result, immunizations that initially provided 90% protection against mild episodes of the disease may only provide 70% protection after 6 or 7 months. The percentage of people that need to be vaccinated to attain COVID-19 herd immunity is still being disputed.
Machine Learning Applications in Finance
(2021-04-28) Kodru, Sushma Sree
Now is an opportune time to for everyone. Let us acknowledge that now because it is very important that the current credit scoring system in place needs to be improved and accurate to avoid any risk going forward. We know for sure that Artificial Intelligence can help in the credit scoring systems process

Browse

Recent Submissions