Computer Engineering Commons^™

3d Convolutional Neural Networks For The Diagnosis Of 6 Unique Pathologies On Head Ct, Travis Clarke, Paras Lakhani, Md

Phase 1

Introduction: Head CT scans are a standard first-line tool used by physicians in the diagnosis of neurological pathologies. Recently, the development of deep learning models such as convolutional neural networks (CNNs) has allowed the rapid identification of bleeds and other pathologies on CT scans. This study aims to show that by training 3D CNNs with a larger, curated dataset, a more comprehensive list of potential diagnoses can be included in the detailed model.

Methods: A retrospective study was performed using a dataset of 66,000 head CT studies from the Thomas Jefferson University health system. Studies were acquired using a natural …

Distributed Multi-Agent Optimization And Control With Applications In Smart Grid, Towfiq Rahman

Electronic Theses and Dissertations, 2020-2023

With recent advancements in network technologies like 5G and Internet of Things (IoT), the size and complexity of networked interconnected agents have increased rapidly. Although centralized schemes have simpler algorithm design, in practicality, it creates high computational complexity and requires high bandwidth for centralized data pooling. In this dissertation, for distributed optimization of networked multi-agent architecture, the Alternating Direction Method of Multipliers (ADMM) is investigated. In particular, a new adaptive-gain ADMM algorithm is derived in closed form and under the standard convex property to greatly speed up the convergence of ADMM-based distributed optimization. Using the Lyapunov direct approach, the proposed …

Data-Driven Nonlinear Control Designs For Constrained Systems, Roland Harvey

Electronic Theses and Dissertations, 2020-2023

Systems with nonlinear dynamics are theoretically constrained to the realm of nonlinear analysis and design, while explicit constraints are expressed as equalities or inequalities of state, input, and output vectors of differential equations. Few control designs exist for systems with such explicit constraints, and no generalized solution has been provided. This dissertation presents general techniques to design stabilizing controls for a specific class of nonlinear systems with constraints on input and output, and verifies that such designs are straightforward to implement in selected applications. Additionally, a closed-form technique for an open-loop problem with unsolvable dynamic equations is developed. Typical optimal …

Enabling Recovery Of Secure Non-Volatile Memories, Mao Ye

Electronic Theses and Dissertations, 2020-2023

Emerging non-volatile memories (NVMs), such as phase change memory (PCM), spin-transfer torque RAM (STT-RAM) and resistive RAM (ReRAM), have dual memory-storage characteristics and, therefore, are strong candidates to replace or augment current DRAM and secondary storage devices. The newly released Intel 3D XPoint persistent memory and Optane SSD series have shown promising features. However, when these new devices are exposed to events such as power loss, many issues arise when data recovery is expected. In this dissertation, I devised multiple schemes to enable secure data recovery for emerging NVM technologies when memory encryption is used. With the data-remanence feature of …

Exploration Of Approaches To Arabic Named Entity Recognition, Husamelddin Balla, Sarah Jane Delany

Conference papers

Abstract. The Named Entity Recognition (NER) task has attracted significant attention in Natural Language Processing (NLP) as it can enhance the performance of many NLP applications. In this paper, we compare English NER with Arabic NER in an experimental way to investigate the impact of using different classifiers and sets of features including language-independent and language-specific features. We explore the features and classifiers on five different datasets. We compare deep neural network architectures for NER with more traditional machine learning approaches to NER. We discover that most of the techniques and features used for English NER perform well on Arabic …

Lung Cancer Subtype Differentiation From Positron Emission Tomography Images, Oğuzhan Ayyildiz, Zafer Aydin, Bülent Yilmaz, Seyhan Karaçavuş, Kübra Şenkaya, Semra İçer, Erdem Arzu Taşdemi̇r, Eser Kaya

Turkish Journal of Electrical Engineering and Computer Sciences

Lung cancer is one of the deadly cancer types, and almost 85 % of lung cancers are nonsmall cell lung cancer (NSCLC). In the present study we investigated classification and feature selection methods for the differentiation of two subtypes of NSCLC, namely adenocarcinoma (ADC) and squamous cell carcinoma (SqCC). The major advances in understanding the effects of therapy agents suggest that future targeted therapies will be increasingly subtype specific. We obtained positron emission tomography (PET) images of 93 patients with NSCLC, 39 of which had ADC while the rest had SqCC. Random walk segmentation was applied to delineate three-dimensional tumor …

On The Automorphisms And Isomorphisms Of Mds Matrices And Their Efficient Implementations, Muharrem Tolga Sakalli, Sedat Akleylek, Kemal Akkanat, Vincent Rijmen

Turkish Journal of Electrical Engineering and Computer Sciences

In this paper, we explicitly define the automorphisms of MDS matrices over the same binary extension field. By extending this idea, we present the isomorphisms between MDS matrices over $\mathbb{F}_{2^{m}}$ and MDS matrices over $\mathbb{F}_{2^{mt}}$, where $t \ge 1$ and $m>1$, which preserves the software implementation properties in view of XOR operations and table lookups of any given MDS matrix over $\mathbb{F}_{2^{m}}$. Then we propose a novel method to obtain distinct functions related to these automorphisms and isomorphisms to be used in generating isomorphic MDS matrices (new MDS matrices in view of implementation properties) using the existing ones. The …

Wideband Patch Array Antenna Using Superstrate Configuration For Future 5gapplications, Sidra Farhat, Farzana Arshad, Yasar Amin, Jonathan Loo

Turkish Journal of Electrical Engineering and Computer Sciences

In this work, four distinct antenna configurations for future-centric 5G applications are proposed. Initially, a single rectangular patch is designed to operate at the frequency of 28 GHz while maintaining a wide operational band. Performance of the antenna is improved by incorporating an array of identical rectangular elements resulting in a higher gain and wider bandwidth. The proposed arrangement consists of three rectangular elements realized using 0.508-mm thick Rogers RT/Duroid 5880 laminate. The bandwidth is further enhanced by increasing the number of radiating elements in the array from three to five. Evolution of the proposed design is concluded by stacking …

Revised Polyhedral Conic Functions Algorithm For Supervised Classification, Gürhan Ceylan, Gürkan Öztürk

Turkish Journal of Electrical Engineering and Computer Sciences

In supervised classification, obtaining nonlinear separating functions from an algorithm is crucial for prediction accuracy. This paper analyzes the polyhedral conic functions (PCF) algorithm that generates nonlinear separating functions by only solving simple subproblems. Then, a revised version of the algorithm is developed that achieves better generalization and fast training while maintaining the simplicity and high prediction accuracy of the original PCF algorithm. This is accomplished by making the following modifications to the subproblem: extension of the objective function with a regularization term, relaxation of a hard constraint set and introduction of a new error term. Experimental results show that …

Cdc42-Dependent Transfer Of Mir301 From Breast Cancer-Derived Extracellular Vesicles Regulates The Matrix Modulating Ability Of Astrocytes At The Blood–Brain Barrier, Golnaz Morad, Cassandra C. Daisy, Hasan H. Otu, Towia A. Libermann, Simon T. Dillon, Marsha A. Moses

Department of Electrical and Computer Engineering: Faculty Publications

Breast cancer brain metastasis is a major clinical challenge and is associated with a dismal prognosis. Understanding the mechanisms underlying the early stages of brain metastasis can provide opportunities to develop efficient diagnostics and therapeutics for this significant clinical challenge. We have previously reported that breast cancer-derived extracellular vesicles (EVs) breach the blood–brain barrier (BBB) via transcytosis and can promote brain metastasis. Here, we elucidate the functional consequences of EV transport across the BBB. We demonstrate that brain metastasis-promoting EVs can be internalized by astrocytes and modulate the behavior of these cells to promote extracellular matrix remodeling in vivo. We …

Power-Over-Tether Uas Leveraged For Nearly Indefinite Meteorological Data Acquisition In The Platte River Basin, Daniel Rico, Carrick Detweiler, Francisco Munoz-Arriola

School of Computing: Conference and Workshop Papers

The integration of unmanned aerial systems (UASs) has increased in the field of agriculture. These systems can provide data that was previously difficult to obtain to help increase efficiency and production. Typical commercial off the shelf (COTS) UASs have significant limitations in the form of small payloads, and short flight times which inhibit their ability to provide significant quantities of useful data. We present the development of a novel power-over-tether UAS that leverages the physical presence of the tether to integrate sensors at multiple altitudes along the tether. The UAS can acquire data nearly indefinitely to sense atmospheric conditions and …

Automated Assessment Of Cardiothoracic Ratios On Chest Radiographs Using Deep Learning, Varun Danda, Paras Lakhani, Md

Phase 1

Introduction: The cardiothoracic ratio (CTR) is a quantitative measure of cardiac size that can measured from chest radiography (CXR). Although radiologists using digital workstations possess the ability to calculate CTR, clinical demands prevent calculation for every case. In this study, the efficacy of a deep convolutional neural network (dCNN) to assess CTR was evaluated.

Methods: 611 HIPAA-compliant de-identified CXRs were obtained from [institution blinded] and public databases. Using ImageJ, a board-certified radiologist (reader #1) and a medical student (reader #2), measured the CTR by marking four pixels on all CXRs: the right- and left-most chest wall, the right- and left-most …

The Influence Of Blockchain Technology On Fraud And Fake Protection, Youngju Yun

OUR Journal: ODU Undergraduate Research Journal

No abstract provided.

Hierarchical Anomaly Detection For Time Series Data, Ryan E. Sperl

Browse all Theses and Dissertations

With the rise of Big Data and the Internet of Things, there is an increasing availability of large volumes of real-time streaming data. Unusual occurrences in the underlying system will be reflected in these streams, but any human analysis will quickly become out of date. There is a need for automatic analysis of streaming data capable of identifying these anomalous behaviors as they occur, to give ample time to react. In order to handle many high-velocity data streams, detectors must minimize the processing requirements per value. In this thesis, we have developed a novel anomaly detection method which makes use …

Corrections To ‘‘Glaciernet: A Deep-Learning Approach For Debris-Covered Glacier Mapping’’, Zhiyuan Xie, Umesh K. Haritashya, Vijayan K. Asari, Brennan W. Young, Michael P. Bishop, Jeffrey S. Kargel

Electrical and Computer Engineering Faculty Publications

In the above article [1], Figure 2 was incorrect. Unfortunately, we mixed the color label of "CONV $\to $ BN $\to $ ReLu" and "Unpooling" in the CNN structure section of Figure 2. The color label of "CONV $\to $ BN $\to $ ReLu" should be orange while the color label of "Unpooling" should be green. Also, the word "Decoder" is misspelled. That same figure with the same error is also used for the graphic abstract. The corrected figure is given here. None of the sections in the figure is modified. The only change is in the color label of …

Applying Artificial Intelligence To Medical Data, Shaikh Shiam Rahman

College of Graduate Studies: Theses & Dissertations

Machine learning, data mining, and deep learning has become the methodology of choice for analyzing medical data and images. In this study, we implemented three different machine learning techniques to medical data and image analysis. Our first study was to implement different log base entropy for a decision tree algorithm. Our results suggested that using a higher log base for the dataset with mostly categorical attributes with three or more categories for each attribute can obtain a higher accuracy. For the second study, we analyzed mental health data tuning the parameters of the decision tree (splitting method, depth and entropy). …

Quantitative Performance Assessment Of Lidar-Based Vehicle Contour Estimation Algorithms For Integrated Vehicle Safety Applications, David M. Mothershed

College of Graduate Studies: Theses & Dissertations

Many nations and organizations are committing to achieving the goal of `Vision Zero' and eliminate road traffic related deaths around the world. Industry continues to develop integrated safety systems to make vehicles safer, smarter and more capable in safety critical scenarios. Passive safety systems are now focusing on pre-crash deployment of restraint systems to better protect vehicle passengers. Current commonly used bounding box methods for shape estimation of crash partners lack the fidelity required for edge case collision detection and advanced crash modeling. This research presents a novel algorithm for robust and accurate contour estimation of opposing vehicles. The presented …

การสกัดตารางและรายการบนเว็บเป็นอาร์ดีเอฟ, จุลเทพ นันทขว้าง

Chulalongkorn University Theses and Dissertations (Chula ETD)

ทุกวันนี้ ลิงก์เดต้าได้เติบโตเพิ่มขึ้นอย่างรวดเร็วตามการเติบโตของเว็บ นอกเหนือจากข้อมูลใหม่ที่สร้างขึ้นในรูปแบบซีแมนติกโดยเฉพาะ ส่วนหนึ่งมาจากการแปลงข้อมูลโครงสร้างที่มีอยู่ให้อยู่ในรูปแบบของข้อมูลเปิดระดับห้าดาว อย่างไรก็ตามยังคงมีข้อมูลจำนวนมากในรูปแบบโครงสร้างและกึ่งโครงสร้าง ตัวอย่างเช่นตารางและรายการซึ่งเป็นรูปแบบหลักที่มนุษย์ใช้อ่าน ยังรอการแปลงอยู่ งานวิจัยนี้กล่าวถึงงานวิจัยต่าง ๆ ที่เกี่ยวกับการแปลงตารางและรายการมาเป็นข้อมูลในรูปแบบต่าง ๆ เพื่อให้เครื่องสามารถอ่านได้ นอกจากนี้ยังเสนอวิธีการในการแปลงตารางและรายการเป็นรูปแบบ Resource Description Framework และยังคงเก็บโครงสร้างต้นฉบับที่จำเป็นไว้อย่างละเอียด ซึ่งทำให้สามารถที่จะสร้างข้อมูลโครงสร้างเดิมกลับมาได้ ระบบ TULIP ถูกสร้างขึ้นเพื่อเป็นเครื่องมือสำหรับการพัฒนาซีแมนติกเว็บ วิธีการที่เสนอมีความยืดหยุ่นมากกว่าเมื่อเทียบกับงานอื่น ๆ เดต้าโมเดลของ TULIP สามารถรองรับการเก็บข้อมูลต้นฉบับอย่างครบถ้วน และสามารถนำมาแสดงใหม่ในมุมมองที่แตกต่างไปจากเดิม เครื่องมือนี้สามารถใช้สร้างข้อมูลจำนวนมหาศาลสำหรับเครื่องคอมพิวเตอร์เพื่อให้ใช้งานได้กว้างมากขึ้นกว่าเดิม

การสรุปใจความสำคัญของข้อความแบบสกัดสำหรับข่าวท่องเที่ยวภาษาไทย, ศรัญญา นาทองห่อ

Chulalongkorn University Theses and Dissertations (Chula ETD)

ปัจจุบันเทคโนโลยีทางด้านคอมพิวเตอร์มีความสำคัญต่อการดำเนินชีวิตประจำวันของมนุษย์เป็นอย่างมากและยังถือว่าเป็นเครื่องมือที่ใช้ในการอำนวยความสะดวกให้แก่มนุษย์มากมายโดยเฉพาะทางด้านการสื่อสารผ่านสังคมออนไลน์ เพื่อลดเวลาในการอ่านข่าวหรืออ่านบทความและข่าวออนไลน์ต่างๆ จากการวิจัยที่ผ่านมามีการศึกษาและพัฒนาการสรุปใจความสำคัญของภาษาไทยเป็นจำนวนมาก ในงานวิจัยนี้ได้นำเสนอวิธีการสรุปใจความสำคัญจากข่าวการท่องเที่ยวภาษาไทย 2 วิธีคือการเลือกประโยคจากการจัดกลุ่มประโยคด้วยเคมีนและการเลือกประโยคด้วยวิธีหาคำสำคัญประโยคจากหัวข้อข่าว โดยมีการพัฒนาและสร้างคลังข้อมูลรายการคำประสมเพื่อช่วยเพิ่มประสิทธิภาพในการตัดคำ โดยการทดลองนี้ใช้ข้อมูลข่าวการท่องเที่ยวไทย ทั้งหมด 400 ข่าวสำหรับใช้ทดลองในการสรุปใจความสำคัญ และ 5,000 ข่าวสำหรับการสร้างคลังข้อมูลรายการคำประสม การวัดประสิทธิภาพของวิธีการที่นำเสนอ มีการวัดประสิทธิภาพการสรุปใจความสำคัญโดยการเปรียบเทียบผลจากการสรุปที่ได้จากผู้เชี่ยวชาญด้านภาษาไทยเทียบกับผลสรุปที่ได้จากวิธีการที่นำเสนอ จากงานวิจัยนี้ในขั้นตอนการสร้างคำประสมได้คำประสมทั้งหมด จำนวน 2,340 คำ ผลการทดลองพบว่าวิธีตัดคำด้วยคัตคำร่วมกับตัดคำประสมได้ผลดีกว่าการตัดคำจากคัตคำเพียงอย่างเดียว และการสรุปใจความสำคัญโดยใช้การคำนวณค่าน้ำหนักของคำสำคัญโดยหาค่าความถี่ของคำจากหัวข้อข่าวเพียงอย่างเดียวและเลือกประโยคเรียงลำดับจากผลรวมความถี่ของคำสำคัญจากหัวข้อข่าวมีประสิทธิภาพและความแม่นยำสูงสุดโดยมีค่าความแม่นยำ ค่าความระลึกและค่าวัดประสิทธิภาพอยู่ที่ 0.8097 0.8367 และ 0.8216 ตามลำดับและเมื่อใช้คัตคำร่วมกับการตัดคำแบบเอ็นแกรมโดยวิธีการสรุปใจความสำคัญแบบเดียวกันได้ค่าความแม่นยำ ค่าความระลึกและค่าวัดประสิทธิภาพอยู่ที่ 0.8119 0.8398 และ 0.8242 ตามลำดับที่อัตราการบีบอัดร้อยละ 20

Semantic Segmentation Using Modified U-Net Architecture For Crack Detection, Michael Sun

Electronic Theses and Dissertations

The visual inspection of a concrete crack is essential to maintaining its good condition during the service life of the bridge. The visual inspection has been done manually by inspectors, but unfortunately, the results are subjective. On the other hand, automated visual inspection approaches are faster and less subjective. Concrete crack is an important deficiency type that is assessed by inspectors. Recently, various Convolutional Neural Networks (CNNs) have become a prominent strategy to spot concrete cracks mechanically. The CNNs outperforms the traditional image processing approaches in accuracy for the high-level recognition task. Of them, U-Net, a CNN based semantic segmentation …

Kernel-Controlled Dqn Based Cnn Pruning For Model Compression And Acceleration, Romancha Khatri

Electronic Theses and Dissertations

Apart from the accuracy, the size of convolutional neural networks (CNN) models is another principal factor for facilitating the deployment of models on memory, power and budget constrained devices. However, conventional model compression techniques require human experts to setup parameters to explore the design space which is suboptimal and time consuming. Various pruning techniques are implemented to gain compression, trading off speed and accuracy. Given a CNN model [11], we propose an automated deep reinforcement learning [9] based model compression technique that can effectively turned off kernels on each layer by observing its significance on decision making. By observing accuracy, …

Determinants Of Startup Funding: The Interaction Between Web Attention And Culture, Jie Ren, Viju Raghupathi, Wullianallur Raghupathi

Journal of International Technology and Information Management

Technology empowers entrepreneurs to pursue alternative funding through platforms like crowdfunding. This research explores significant startup funding factors using Crunchbase. Controlling for common factors (acquisition/funding-rounds/IPO), the research uniquely focuses on web attention - the visibility on social media - and its impact on funding. It also examines the moderating influence of startup’s home country culture (individualism/collectivism). Findings show stronger positive impact of web attention on startup funding for collectivist countries. While individualistic investors value personal goals, collectivists value collaborative goals - inclinations that align with crowdfunding behavior. Therefore while increasing web attention, crowdfunding efforts can be targeted towards collectivist countries.

Food Printing: Evolving Technologies, Challenges, Opportunities, And Best Adoption Strategies, Sharmin Attarin, Mohsen Attaran

Journal of International Technology and Information Management

3D printing is a process of making three-dimensional objects using additive processes where layers are laid down in succession to create a complete object. Companies across the globe are actively piloting and leveraging the inherent benefits of 3D printing technology. Today, 3D printing can revolutionize food innovation and production through better creativity, customizability, and sustainability. This article highlights 3D printing evolving technologies and trends and identifies its applications and implementation challenges. In this paper, we conducted a literature review to explore 3D printing's current technologies and applications in the food industry, including its advantages, its potential implications, and its obstacles …

Evaluating Students Information System Success Using Delone And Mclean’S Model: Student’S Perspective, Majaliwa Mkinga, Herman Mandari

Journal of International Technology and Information Management

System success is considered to be an important element in accomplishing the goals of the organization; therefore evaluation of system success needs to be done in order to ensure that investment in Information System is successful. Most of Higher Learning Institutions (HLIs) in Tanzania have adopted the use of IS in providing service to their customers. Nevertheless, there is less evidence that system success evaluation has been done in order to identify the desired characteristics which could make IS more effective. Due to that, this study evaluates the effectiveness of Student Information System (SIS) used at the Institute of Finance …

An Integrated View Of Data: Application Of Knowledge Modeling To Data Management, Sung-Kwan Kim, Wenjun Wang

Journal of International Technology and Information Management

Data management has become an important challenge. Good data management requires an effective approach to collecting, storing, and accessing data across the enterprise. In this paper, a knowledge modeling approach to data management is introduced with an emphasis on data requirements analysis. A knowledge model can provide a high-level view of organizational data by specifying the structure and relationships of the knowledge contents used in business processes. The proposed knowledge modeling approach is business process oriented and decision oriented. The description of the knowledge contents in the model is based on ontological specification. The model is comprised of five elements: …

Uga’S Alexander Campbell King Law Library: Phasing In Inclusive Usability Testing, Rachel S. Evans, Marie Mize, Jason Tubinis

Articles, Chapters and Online Publications

For years we have offered our EBSCO discovery layer service (EDS) as a secondary search tool in addition to our traditional online catalog (GAVEL) linking to both from the library website. However, the traditional catalog search, also known as “Classic GAVEL”, is always listed first while EDS, also known as “GAVEL & Beyond”, is listed second. Although maintenance has continued for populating EDS with library records on a daily basis, customization for this interface and sharing it with our users has not been prioritized. Before making any decisions related to changing the primary location our users experience when searching the …

Leveraging Peer-To-Peer Energy Sharing For Resource Optimization In Mobile Social Networks, Aashish Dhungana

Theses and Dissertations

Mobile Opportunistic Networks (MSNs) enable the interaction of mobile users in the vicinity through various short-range wireless communication technologies (e.g., Bluetooth, WiFi) and let them discover and exchange information directly or in ad hoc manner. Despite their promise to enable many exciting applications, limited battery capacity of mobile devices has become the biggest impediment to these appli- cations. The recent breakthroughs in the areas of wireless power transfer (WPT) and rechargeable lithium batteries promise the use of peer-to-peer (P2P) energy sharing (i.e., the transfer of energy from the battery of one member of the mobile network to the battery of …

Benefits, Drawbacks, And Risks Of Ai, James M. Donovan

Law Faculty Books and Chapters

The specter of the impact of artificial intelligence [AI] on law and legal education casts a long and uncertain shadow. Its mix of enthusiasm and trepidation can arise from a thin idea of what the label refers to. Four components differentiate AI from even high-end automation: Big data and predictive analytics, deep learning software, cloud computing, and natural language process. From the perspective of the person on the street though, is captured as “the art of creating machines that perform functions that require intelligence when performed by people,” centering on the ability to make independent choices. While that gloss may …

Cooperative Communications With Optimal Harvesting Duration For Nakagamifading Channels, Nadhir Ben Halima, Boujemaa Hatem

Turkish Journal of Electrical Engineering and Computer Sciences

In this paper, we analyze the throughput of cooperative communications with wireless energy harvesting. Relay nodes harvest energy from Radio Frequency (RF) signal transmitted by the source. We derive the packet error probability as well as the throughput for Nakagami fading channels. We also suggest to enhance the throughput by choosing the value of harvesting duration. Our results are valid for both Amplify and Forward (AF) and Decode and Forward (DF) relaying.

Two Novel Radar Detectors For Spiky Sea Clutter With The Presence Of Thermal Noise And Interfering Targets, Nouh Guidoum, Faouzi Soltani, Amar Mezache

Turkish Journal of Electrical Engineering and Computer Sciences

In the context of noncoherent detection and high-resolution maritime radar system with low grazing angle, new Constant False Alarm Rate (CFAR) decision rules are suggested for two Compound Gaussian (CG) clutters namely: The K distribution and the Compound Inverse Gaussian (CIG) distribution, which are considered among the most appropriate models for sea clutter. The proposed decision rules are then modified to deal with the presence of thermal noise and interfering targets. The proposed detectors are investigated on the basis of synthetic data as well as real data of the IPIX radar database. The obtained results exhibit a high probability of …