Computer Sciences Commons^™

On Explainability Of Neural Networks, Cem Benar

Dissertations

It is widely reported that deep neural networks outperform most competitors for a range of applications. The state-of-the-art neural networks have built-in inductive bias of architectural choices, regularizations, optimizer types, and initialization methods. Using inductive bias is intuitive to enhance the model approximation. Deep neural networks are mostly dense and heavily overparameterized. They tend to be biased towards low-rank solutions to reduce complexity and improve generalization performance, known as implicit regularization. The implicit regularization as observed in specific architectures and various real-world data sets suggests to overparameterize neural networks judiciously and learn compressed representations (lower rank approximation) with improved performance. …

Vertical Federated Learning Using Autoencoders With Applications In Electrocardiograms, Wesley William Chorney

Theses and Dissertations

Federated learning is a framework in machine learning that allows for training a model while maintaining data privacy. Moreover, it allows clients with their own data to collaborate in order to build a stronger, shared model. Federated learning is of particular interest to healthcare data, since it is of the utmost importance to respect patient privacy while still building useful diagnostic tools. However, healthcare data can be complicated — data format might differ across providers, leading to unexpected inputs and incompatibility between different providers. For example, electrocardiograms might differ in sampling rate or number of leads used, meaning that a …

Towards A Robust Defense: A Multifaceted Approach To The Detection And Mitigation Of Neural Backdoor Attacks Through Feature Space Exploration And Analysis, Liuwan Zhu

Electrical & Computer Engineering Theses & Dissertations

From voice assistants to self-driving vehicles, machine learning(ML), especially deep learning, revolutionizes the way we work and live, through the wide adoption in a broad range of applications. Unfortunately, this widespread use makes deep learning-based systems a desirable target for cyberattacks, such as generating adversarial examples to fool a deep learning system to make wrong decisions. In particular, many recent studies have revealed that attackers can corrupt the training of a deep learning model, e.g., through data poisoning, or distribute a deep learning model they created with “backdoors” planted, e.g., distributed as part of a software library, so that the …

Generalizable Deep-Learning-Based Wireless Indoor Localization, Ali Owfi

All Theses

The growing interest in indoor localization has been driven by its wide range of applications in areas such as smart homes, industrial automation, and healthcare. With the increasing reliance on wireless devices for location-based services, accurate estimation of device positions within indoor environments has become crucial. Deep learning approaches have shown promise in leveraging wireless parameters like Channel State Information (CSI) and Received Signal Strength Indicator (RSSI) to achieve precise localization. However, despite their success in achieving high accuracy, these deep learning models suffer from limited generalizability, making them unsuitable for deployment in new or dynamic environments without retraining. To …

Well-Conditioned T-Matrix Formulation For Scattering By A Dielectric Obstacle, Murat Enes Hati̇poğlu, Fati̇h Di̇kmen

Turkish Journal of Electrical Engineering and Computer Sciences

The classic formulation of the extended boundary condition method is revisited to inject the regularization operators for the unknown coefficients of the eigen-function expansions for the travelling and standing waves throughout the dielectric scatterer. It is shown that, using the new definitions, the existing algorithm of the scattering field calculation can be kept the same for its well-conditioned version. This is exemplified for scalar 2D problems for both TM and TE polarization under illumination of a line source. The condition numbers of the matrix operators in the new version of the algorithm are drastically reduced when the regularization interfaces are …

Characterization And Estimation Of Musculoskeletal Pain Using Machine Learning, Boluwatife Faremi

Master's Theses

Traditional scales utilized for recording pain are known to be highly subjective and biased due to inaccuracies in recollecting actual pain intensities. As a result, machine learning (ML) models that are trained using these scores as ground truth are reported to have low performance for objective pain classification because of the huge disparity between what was felt in moments of pain and the scores recorded afterward.

In the present study, two devices were designed for gathering real-time, continuous in-session subjective pain scores and the recording of the autonomic nervous system (ANS) altered endodermal (EDA) activity. 24 participants were recruited to …

A Comparative Effectiveness Study On Opioid Use Disorder Prediction Using Artificial Intelligence And Existing Risk Models, Sajjad Fouladvand, Jeffery Talbert, Linda Phyliss Dwoskin, Heather M. Bush, Amy L. Meadows, Lars E. Peterson, Yash R. Mishra, Steven K. Roggenkamp, Fei Wang, Ramakanth Kavuluru, Jin Chen

Markey Cancer Center Faculty Publications

Opioid use disorder (OUD) is a leading cause of death in the United States placing a tremendous burden on patients, their families, and health care systems. Artificial intelligence (AI) can be harnessed with available healthcare data to produce automated OUD prediction tools. In this retrospective study, we developed AI based models for OUD prediction and showed that AI can predict OUD more effectively than existing clinical tools including the unweighted opioid risk tool (ORT). Data include 474,208 patients’ data over 10 years; 269,748 were females with an average age of 56.78 years. Cases are prescription opioid users with at least …

Lightweight Deep Neural Network Models For Electromyography Signal Recognition For Prosthetic Control, Ahmet Mert

Turkish Journal of Electrical Engineering and Computer Sciences

In this paper, lightweight deep learning methods are proposed to recognize multichannel electromyography (EMG) signals against varying contraction levels. The classical machine learning, and signal processing methods namely, linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), root mean square (RMS), and waveform length (WL) are adopted to convolutional neural network (CNN), and long short-term memory neural network (LSTM). Eight-channel recordings of nine amputees from a publicly available dataset are used for training and testing the proposed models considering prosthetic control strategies. Six class hand movements with three contraction levels are applied to WL and RMS-based feature extraction. After that, they …

A Practical Framework For Early Detection Of Diabetes Using Ensemble Machine Learning Models, Qusay Saihood, Emrullah Sonuç

Turkish Journal of Electrical Engineering and Computer Sciences

The diagnosis of diabetes, a prevalent global health condition, is crucial for preventing severe complications. In recent years, there has been a growing effort to develop intelligent diagnostic systems for diabetes utilizing machine learning (ML) algorithms. Despite these efforts, achieving high accuracy rates using such systems remains a significant challenge. Recent advancements in ensemble ML methods offer promising opportunities for early detection of diabetes, as they are known to be faster and more cost-effective than traditional approaches. Therefore, this study proposes a practical framework for diagnosing diabetes that involves three stages. The data preprocessing stage encompasses several crucial tasks, including …

Improving Unet Segmentation Performance Using An Ensemble Model In Images Containing Railway Lines, Mehmet Sevi̇, İlhan Aydin

Turkish Journal of Electrical Engineering and Computer Sciences

This study aims to make sense of the autonomous system and the railway environment for railway vehicles. For this purpose, by determining the railway line, information about the general condition of the line can be obtained along the way. In addition, objects such as pedestrian crossings, people, cars, and traffic signs on the line will be extracted. The rails and the rail environment in the images will be segmented with a semantic segmentation network. In order to ensure the safety of rail transport, computer vision, and deep learning-based methods are increasingly used to inspect railway tracks and surrounding objects. In …

Optimization Techniques For Machine Learning Inference And Near Memory Image Processing In Hardware For Highly Constrained Iot Edge Nodes, Rajeev Joshi

USF Tampa Graduate Theses and Dissertations

The growing demand for fast and energy-efficient hardware for resource-constrained Internet of Things (IoT) edge devices has highlighted the limitations of conventional computing architectures. This research focuses on addressing the demand for fast, optimized, and energy-efficient machine learning inference engines as well as image processing in IoT edge applications. In this work, we address three challenging research problems and devise efficient solutions. Our investigation involved comprehensive exploration and analysis, leading to the proposal of effective approaches for overcoming these demanding issues. Through our work, we contribute novel solutions that offer improved efficiency and effectiveness in handling these research problems.

First, …

On The Use Of Wireless Technologies For Wildlife Monitoring: Wireless Sensor Network Routing Protocols, Lilian Mutalemwa

Tanzania Journal of Engineering and Technology (TJET)

Traditional methods for wildlife monitoring are labor-intensive and time-consuming. Therefore, advanced technologies and remote monitoring methods are becoming increasingly popular. This paper presents a study on wireless technologies for wildlife monitoring. In the study, a review of the literature was done to identify the most commonly used wireless technologies. Various technologies were explored including unmanned aerial vehicles (UAVs), Internet of Things (IoT), wireless sensor networks (WSNs), artificial intelligence (AI), global positioning system (GPS), and very high frequency (VHF) radio. Then, a more detailed study was done on WSN technology. Investigations were done to observe the performance of routing protocols in …

Deep Learning Enhancement And Privacy-Preserving Deep Learning: A Data-Centric Approach, Hung S. Nguyen

USF Tampa Graduate Theses and Dissertations

Deep Learning and its applications have become attractive to a lot of research recentlybecause of its capability to capture important information from large amounts of data. While most of the work focuses on finding the best model parameters, improving machine learning performance from data perspective still needs more attention. In this work, we propose techniques to enhance the robustness of deep learning classification by tackling data issue. Specifically, our data processing proposals aim to alleviate the impacts of class-imbalanced data and non- IID data in deep learning classification and federated learning scenarios. In addition, data pre-processing strategies such that dimensionality …

Adversary Aware Continual Learning, Muhammad Umer

Theses and Dissertations

Continual learning approaches are useful as they help the model to learn new information (classes) sequentially, while also retaining the previously acquired information (classes). However, these approaches are adversary agnostic, i.e., they do not consider the possibility of malicious attacks. In this dissertation, we have demonstrated that continual learning approaches are extremely vulnerable to the adversarial backdoor attacks, where an intelligent adversary can introduce small amount of misinformation to the model in the form of imperceptible backdoor pattern during training to cause deliberate forgetting of a specific class at test time. We then propose a novel defensive framework to counter …

Novel Approach For Non-Invasive Prediction Of Body Shape And Habitus, Emma Young

Electronic Theses and Dissertations

While marker-based motion capture remains the gold standard in measuring human movement, accuracy is influenced by soft-tissue artifacts, particularly for subjects with high body mass index (BMI) where markers are not placed close to the underlying bone. Obesity influences joint loads and motion patterns, and BMI may not be sufficient to capture the distribution of a subject’s weight or to differentiate differences between subjects. Subjects in need of a joint replacement are more likely to have mobility issues or pain, which prevents exercise. Obesity also increases the likelihood of needing a total joint replacement. Accurate movement data for subjects with …

Neuroevolution Application To Collaborative And Heuristics-Based Connected And Autonomous Vehicle Cohort Simulation At Uncontrolled Intersection, Frederic Jacquelin, Jungyun Bae, Bo Chen, Darrell Robinette

Michigan Tech Publications

Artificial intelligence is gaining tremendous attractiveness and showing great success in solving various problems, such as simplifying optimal control derivation. This work focuses on the application of Neuroevolution to the control of Connected and Autonomous Vehicle (CAV) cohorts operating at uncontrolled intersections. The proposed method implementation’s simplicity, thanks to the inclusion of heuristics and effective real-time performance are demonstrated. The resulting architecture achieves nearly ideal operating conditions in keeping the average speeds close to the speed limit. It achieves twice as high mean speed throughput as a controlled intersection, hence enabling lower travel time and mitigating energy inefficiencies from stop-and-go …

Eddy Current Defect Response Analysis Using Sum Of Gaussian Methods, James William Earnest

Theses and Dissertations

This dissertation is a study of methods to automatedly detect and produce approximations of eddy current differential coil defect signatures in terms of a summed collection of Gaussian functions (SoG). Datasets consisting of varying material, defect size, inspection frequency, and coil diameter were investigated. Dimensionally reduced representations of the defect responses were obtained utilizing common existing reduction methods and novel enhancements to them utilizing SoG Representations. Efficacy of the SoG enhanced representations were studied utilizing common Machine Learning (ML) interpretable classifier designs with the SoG representations indicating significant improvement of common analysis metrics.

Secure And Efficient Federated Learning, Xingyu Li

Theses and Dissertations

In the past 10 years, the growth of machine learning technology has been significant, largely due to the availability of large datasets for training. However, gathering a sufficient amount of data on a central server can be challenging. Additionally, with the rise of mobile networking and the large amounts of data generated by IoT devices, privacy and security issues have become a concern, resulting in government regulations such as GDPR, HIPAA, CCPA, and ADPPA. Under these circumstances, traditional centralized machine learning methods face a problem in that sensitive data must be kept locally for privacy reasons, making it difficult to …

Uibee: An Improved Deep Instance Segmentation And Classification Of Ui Elements In Wireframes, Cahi̇t Berkay Kazangi̇rler, Caner Özcan, Buse Yaren Teki̇n

Turkish Journal of Electrical Engineering and Computer Sciences

User Interface (UI) is a basic concept in which individuals interact with any computer program or technological device to create a graphical design. In the initial stages of app development, UI prototype is a must. An automatic analysis system for the basic execution of UI designs will considerably speed up the development of designs according to old-fashioned methods. In this approach, it is aimed at saving cost and time by automating the process. For the aforesaid objective, we present a new approach rather than the traditional methods. For this reason, a high amount of elements in wireframes are detected and …

Efficient Modelling Of Random Access Memory Cell: An Approach Using Qca Nanocomputing, Ali Newaz Bahar, Angshuman Khan

Turkish Journal of Electrical Engineering and Computer Sciences

Quantum-dot cellular automata (QCA) is innovative and potentially fruitful nanotechnology that provides a solution for transistor-based circuits with enhanced switching frequency, large-scale integration, and low power consumption. The random-access memory (RAM) cell is a fundamental component that is designed to operate quickly and effectively since memory is a core part of the semiconductor industry, thus the QCA family. The RAM cell design in this work is based on a multiplexer structure and is implemented without using coplanar crossovers of QCA technology. QCADesigner-2.0.3, a standard QCA layout design and verification tool, is used in the simulation and validation processes for the …

Trace Dna Detection Using Diamond Dye: A Recovery Technique To Yield More Dna, Leah Davis

Master's Theses

This study aspires to find a new screening approach to trace DNA recovery techniques to yield a higher quantity of trace DNA from larger items of evidence. It takes the path of visualizing trace DNA on items of evidence with potential DNA so analysts can swab a more localized area rather than attempting to recover trace DNA through the general swabbing technique currently used for trace DNA recovery. The first and second parts consisted of observing trace DNA interaction with Diamond Dye on porous and non-porous surfaces.

The third part involved applying the Diamond Dye solution by spraying it onto …

Task Offloading And Resource Allocation Based On Dl-Ga In Mobile Edge Computing, Hang Gu, Minjuan Zhang, Wenzao Li, Yuwen Pan

Turkish Journal of Electrical Engineering and Computer Sciences

With the rapid development of 5G and the Internet of Things (IoT), the traditional cloud computing architecture struggle to support the booming computation-intensive and latency-sensitive applications. Mobile edge computing (MEC) has emerged as a solution which enables abundant IoT tasks to be offloaded to edge services. However, task offloading and resource allocation remain challenges in MEC framework. In this paper, we add the total number of offloaded tasks to the optimization objective and apply algorithm called Deep Learning Trained by Genetic Algorithm (DL-GA) to maximize the value function, which is defined as a weighted sum of energy consumption, latency, and …

Development Of A Cost-Constrained Intelligent Prosthetic Knee With Real-Time Machine Learning, Predictive Stumble Control, Lucas Jonathan Galey

Open Access Theses & Dissertations

The field of biomechatronics is evolving quickly with advances in computer science, biology, and electrical and mechanical engineering. Coupled with increased interests in machine learning (ML) across all industry sectors, there are opportunities to leverage advanced analytics in uniquely complex problems. This study aimed to deploy real-time ML predictions in a novel microprocessor-controlled prosthetic knee (MPK) device capable of identifying and responding to stumble-events to reduce amputee fall prevalence. Innately, stumbling is a chaotic event. Current MPKs operate by detecting gait characteristics and reacting to preprogrammed states. While these systems are beneficial in significant ways, such as energy expenditure and …

Procedural City Generation With Combined Architectures For Real-Time Visualization, Griffin Poyck

All Theses

The work and research of this paper sought to build upon traditional city generation and simulation in creating a tool that both realistically simulates cities and their prominent features and also creates aesthetic and artistically rich cities using assets that combine several contemporary or near contemporary architectural styles. The major city features simulated are the surrounding terrain, road networks, individual buildings, and building placement. The tools used to both create and integrate these features were created in Houdini with Unreal Engine 5 as the intended final destination. This research was influenced by the city, town, and road networking of Ghost …

Modeling, Simulation And Control Of Microrobots For The Microfactory., Zhong Yang

Electronic Theses and Dissertations

Future assembly technologies will involve higher levels of automation in order to satisfy increased microscale or nanoscale precision requirements. Traditionally, assembly using a top-down robotic approach has been well-studied and applied to the microelectronics and MEMS industries, but less so in nanotechnology. With the boom of nanotechnology since the 1990s, newly designed products with new materials, coatings, and nanoparticles are gradually entering everyone’s lives, while the industry has grown into a billion-dollar volume worldwide. Traditionally, nanotechnology products are assembled using bottom-up methods, such as self-assembly, rather than top-down robotic assembly. This is due to considerations of volume handling of large …

Rattus Norvegicus As A Biological Detector Of Clandestine Remains And The Use Of Ultrasonic Vocalizations As A Locating Mechanism, Gabrielle M. Johnston

Master's Theses

In investigations, locating missing persons and clandestine remains are imperative. One way that first responder and police agencies can search for the remains is by using cadaver dogs as biological detectors. Cadaver dogs are typically used due to their olfactory sensitivity and ability to detect low concentrations of volatile organic compounds produced by biological remains. Cadaver dogs are typically chosen for their stamina, agility, and olfactory sensitivity. However, what is not taken into account often is the size of the animal and the expense of maintaining and training the animal. Cadaver dogs are typically large breeds that cannot fit in …

Analysis And Implementation Of A New High-Buck Dc-Dc Converter With Interleaved Output Inductors And Soft Switching Capability, Sajad Ghabeli Sani, Mohamad Reza Banaei, Seyed Hossein Hosseini

Turkish Journal of Electrical Engineering and Computer Sciences

This paper proposes an innovative structure for DC-DC converters with high buck gain by using a lower number of elements. The converter provides highly efficient output power and an extended output voltage range. In addition, the distribution of output current between two inductors and the soft-switching capability of the power switches have made the converter suitable for applications that require high output current. All power switches accomplish the ZVZCS (zero-voltage and zero-current switching) condition with the aid of a small auxiliary inductor (Lx), which charges and discharges parallel capacitors of main switches to provide soft-switching conditions. Thus, the switching losses …

Integration Of Neural Network And Distance Relay To Improve The Fault Localization On Transmission Lines, Linh Tran

Turkish Journal of Electrical Engineering and Computer Sciences

Power transmission lines are integral and very important components of power systems. Because of the length of these lines and the complexity of the power grids, the lines may encounter various incidents such as lightning strike, shortage, and breakage. When an incident or a fault occurs, a fast process of identification, localization, and isolation of the fault is desired. An accurate fault localization would have a great impact in reducing the restoration time of the system. One of the most popular solutions for fault detection and localization is the distance relays using the impedance-based algorithms. However, these relays are still …

Deep Learning-Based Turkish Spelling Error Detection With A Multi-Class False Positive Reduction Model, Burak Aytan, Cemal Okan Şakar

Turkish Journal of Electrical Engineering and Computer Sciences

Spell checking and correction is an important step in the text normalization process. These tasks are more challenging in agglutinative languages such as Turkish since many words can be derived from the root word by combining many suffixes. In this study, we propose a two-step deep learning-based model for misspelled word detection in the Turkish language. A false positive reduction model is integrated into the system to reduce the false positive predictions originating from the use of foreign words and abbreviations that are commonly used in Internet sharing platforms. For this purpose, we create a multi-class dataset by developing a …

Breath Analysis For Detection Of Lung Cancer With Hybrid Sensor-Based Electronic Nose, Ümi̇t Özsandikcioğlu, Ayten Atasoy

Turkish Journal of Electrical Engineering and Computer Sciences

Lung cancer has the highest death rates among all types of cancer worldwide. Detection of lung cancer in its early stages significantly increases the survival rate. In this study, the aim is to improve the lung cancer detection performance of electronic noses (e-noses) with breath analysis by using two different types of gas sensor-based e-nose. The developed e-nose system consists of 14 quartz crystal microbalance (QCM) sensors and 8 metal oxide semiconductor (MOS) sensors. Breath samples were collected from a total of 100 volunteers, including 60 patients with lung cancer, 20 healthy nonsmokers, and 20 healthy smokers, and were classified …