Digital Commons Network^™

Towards Dynamic Context Detection From Voice Commands And Conversations With Smart Assistants In Smart Homes, Jeniya Sultana

Graduate Theses/Dissertations

Voice-enabled interactions have become increasingly popular with the rise of voice assistants. Identifying contexts or meanings from voice commands and conversations with smart assistants can contribute to the autonomous control of smart home devices and appliances. To improve automation, there is a growing need for efficient context detection that eliminates the need to memorize voice commands. To address this need, I followed a two-step approach in my research. In the first step, I developed a unique context recognition model using a transformer, an attention mechanism, and a fully connected neural network. I trained this model on a conversational dataset of …

Adversarial Transferability And Generalization In Robust Deep Learning, Tao Wu

Doctoral Dissertations

Despite its remarkable achievements across a multitude of benchmark tasks, deep learning (DL) models exhibit significant fragility to adversarial examples, i.e., subtle modifications applied to inputs during testing yet effective in misleading DL models. These meticulously crafted perturbations possess the remarkable property of transferability: an adversarial example that effectively fools one model often retains its effectiveness against another model, even if the two models were trained independently. This research delves into the characteristics influencing the transferability of adversarial examples from three distinct and complementary perspectives: data, model, and optimization. Firstly, from the data perspective, we propose a new method of …

Time Series Anomaly Detection Using Generative Adversarial Networks, Shyam Sundar Saravanan

Masters Theses

"Anomaly detection is widely used in network intrusion detection, autonomous driving, medical diagnosis, credit card frauds, etc. However, several key challenges remain open, such as lack of ground truth labels, presence of complex temporal patterns, and generalizing over different datasets. In this work, we propose TSI-GAN, an unsupervised anomaly detection model for time-series that can learn complex temporal patterns automatically and generalize well, i.e., no need for choosing dataset-specific parameters, making statistical assumptions about underlying data, or changing model architectures. To achieve these goals, we convert each input time-series into a sequence of 2D images using two encoding techniques with …

Learn From The Past: Using Peer Data To Improve Course Recommendations In Personalized Education, Colton Walker

Masters Theses

"This work describes a recommendation approach designed to enhance student success by identifying semester schedules and graduation paths. The primary objective is to provide personalized graduation path recommendations rooted in individual student performance and draw insights from the academic journeys of similar students who successfully graduated. The original research contribution of this work lies in the development of a graduation path recommender system that leverages a combination of Markov Decision Process, Q-Learning, and collaborative filtering techniques to pinpoint graduation paths with a higher likelihood of leading students to success based on their academic progress thus far. The effectiveness of the …

Secure And Privacy-Preserving Federated Learning With Rapid Convergence In Leo Satellite Networks, Mohamed Elmahallawy

Doctoral Dissertations

"The advancement of satellite technology has enabled the launch of small satellites equipped with high-resolution cameras into low Earth orbit (LEO), enabling the collection of extensive Earth data for training AI models. However, the conventional approach of downloading satellite-related data to a ground station (GS) for training a centralized machine learning (ML) model faces significant challenges. Firstly, the transmission of raw data raises security and privacy concerns, especially in military applications. Secondly, the download bandwidth is limited, which puts a stringent limit on image transmissions to the GS. Lastly, LEO satellites have sporadic visibility with the GS, and orbit the …

Identifying Fact-Checks Helpful For Vetting Factual Claims, Theodora Toutountzi

Computer Science and Engineering Dissertations - Archive

Recent efforts to combat misinformation have increasingly focused on automating the fact-checking process. For fact-checking systems to be automated, they need to recognize claims worth checking, match them with previously fact-checked claims, and determine their truthfulness. We refer to the task of matching unvetted claims with previously fact-checked claims as claim-matching. To be precise, the task is defined as follows: given a factual claim, identifying the fact-checks from a repository that could be helpful, or partially helpful, for vetting the given claim. A solution to this task is useful in practice since claimants often repeat the same claims even if …

Efficiently Learning An Encoder That Classifies Token Replacements And Masked Permuted Network-Based Bigru Attention Classifier For Enhancing Sentiment Classification Of Scientific Text, Muhammad Inaam Ul Haq, Khalid Mahmood, Qianmu Li, Ashok Kumar Das, Sachin Shetty, Majid Hussain

VMASC Publications

The exponential growth of scientific literature in digital repositories poses challenges in interpreting complex attitudes within academic texts. Traditional sentiment analysis methods often struggle with nuanced word meanings due to contextual variations. To address this, we propose the Electra-MPNet-based BiGRU attention classifier that extracts the high-level semantic features from citation sentences using the combined strength of Electra and MPNet encoder layers. These features are then combined to extract long-range dependencies through a stacked BiGRU layer. A linear attention mechanism is imposed to estimate the attention weights and context vector which enables the model to selectively focus on relevant information. The …

Multiview Outlier Filtered Pediatric Heart Sound Classification, Sagnik Dakshit

Computer Science Faculty Publications and Presentations

The advancements in deep learning has generated a large-scale interest in development of black-box models for various use cases in different domains such as healthcare, in both at-home and critical setting for diagnosis and monitoring of various health conditions. The use of audio signals as a view for diagnosis is nascent and the success of deep learning models in ingesting multimedia data provides an opportunity for use as a diagnostic medium. For the widespread use of these decision support systems, it is prudent to develop high performing systems which require large quantities of data for training and low-cost method of …

Concept Of Multifactor Method And Non-Functional Requirements Solution To Increase Resilience Through Functional Safety With Cybersecurity Analysis, Emilian Piesik, Marcin Sliwinski, Nary Subramanian, Janusz Zalewski

Computer Science Faculty Publications and Presentations

In the process of designing safety systems, an integrated approach in safety and cybersecurity analysis is necessary. The paper describes a new technique of increasing resilience through integrated analysis of functional safety and cybersecurity. It is a modeling methodology based on the combination of the multifactor method utilizing modified risk graphs, used previously for Safety Integrity Level (SIL) assessment, and the Non-Functional Requirements (NFR) approach. The NFR approach, based on the analysis of graphical representation of conceptual and physical components of the system, contributes a technique to include cybersecurity through the Softgoal Interdependency Graph. The assessment methodology is outlined in …

Fuzzy Convolution Neural Networks For Tabular Data Classification, Arun D. Kulkarni

Computer Science Faculty Publications and Presentations

Recently, convolution neural networks (CNNs) have attracted a great deal of attention due to their remarkable performance in various domains, particularly in image and text classification tasks. However, their application to tabular data classification remains underexplored. There are many fields such as bioinformatics, finance, medicine where non-image data are prevalent. Adaption of CNNs to classify non-image data remains highly challenging. This paper investigates the efficacy of CNNs for tabular data classification, aiming to bridge the gap between traditional machine learning approaches and deep learning techniques. We propose a novel framework fuzzy convolution neural network (FCNN) tailored specifically for tabular data …

The Flesh And Silicon Mesh: A New Materialist Conception Of Advanced Digital Machines, Joshua Nieubuurt

English Faculty Publications

Digital machines, including artificial intelligence (AI), have long been held as potential agents of social change. From their binary role in cultural productions to the real-world implementation of such, digital machines have typically been considered within the Heideggerian “readiness-to-hand” category of Being. This paper seeks to alter conceptually this dynamic through the use of new materialism and Morton’s concept of the “mesh” to provide a new understanding of the social and pragmatic roles advanced digital machines play in ecologies in which they dwell and continue to emerge. Through briefly tracing the complex intermingling of advanced digital machines in the realms …

Demystifying The Hosting Infrastructure Of The Free Content Web: A Security Perspective, Mohammed Alqadhi

Graduate Thesis and Dissertation 2023-2024

This dissertation delves into the security of free content websites, a crucial internet component that presents significant security challenges due to their susceptibility to exploitation by malicious actors. While prior research has highlighted the security disparities between free and premium content websites, it has not delved into the underlying causes. This study aims to address this gap by examining the security infrastructure of free content websites. The research commences with an analysis of the content management systems (CMSs) employed by these websites and their role. Data from 1,562 websites encompassing free and premium categories is collected to identify CMS usage …

Towards Energy-Efficient Edge Computing For Tiny Ai Applications, Vamsi Krishna Bhagavathula

Theses and Dissertations

As artificial intelligence (AI) applications become more common on the edge of networks, like Raspberry Pi servers, it is crucial to optimize their energy use. This research project investigates how AI algorithms affect energy efficiency and resource usage on Raspberry Pi servers. Two models were created: one predicts resource usage, and the other predicts power consumption of AI algorithms on Raspberry Pi. Several factors are considered like CPU and memory use, algorithm speed, dataset size, and types of algorithms and datasets. Using regression-based methods, we model how these factors affect energy use. By converting categorical factors into numerical ones, we …

Blockchain For Computational Integrity And Privacy, Rahul Raj

Theses and Dissertations

This study proposes a blockchain based system that utilizes fully homomorphic encryption to provide security of data in use as well as computational integrity. This is achieved by leveraging the attributes of blockchain which provides availability and data integrity combined with homomorphic encryption that provides confidentiality. The proposed system is designed to perform statistical operations, including mean, median and variance, on encrypted data, thus providing confidentiality of data while in use. The computations are performed on the smart contract, residing on the blockchain which provides computational integrity. The results indicate that it is possible to perform fully homomorphic computations on …

Advancing Explainability In Multi-Label Classification For Tomato Disease Detection Using Machine Learning Interpretability Techniques, Md. Nurullah

Theses and Dissertations

Plant diseases pose a significant threat to global food security, affecting crop yield, quality, and overall agricultural productivity. Traditionally, diagnosing plant diseases has relied on timeconsuming visual inspections by experts, which can often lead to errors. With the rapid growth of technology, machine learning (ML) and artificial intelligence (AI) have opened new possibilities for automating this process. One of the most promising technologies for plant disease diagnosis is Convolutional Neural Networks (CNNs), which have proven effective in image classification tasks. Plant leaves, often exhibiting symptoms such as discoloration and irregular textures, serve as key indicators for disease detection. By processing …

Attitudes And Perceptions Towards Privacy And Surveillance In Australia, Aleatha J. Shanley

Theses: Doctorates and Masters

Understanding attitudes towards privacy and surveillance technologies used to enhance security objectives is a complex, but crucial aspect for policy makers to consider. Historically, terrorism-related incidents justified the uptake of surveillance practices. More recently however, biosecurity concerns have motivated nation-states to adopt more intrusive surveillance measures. There is a growing body of literature that supports the public’s desire to maintain privacy despite fears of biological or physical threats.

This research set out to explore attitudes towards privacy and surveillance in an Australian context. Throughout the course of this endeavour, the COVID-19 pandemic emerged bringing with it a variety of track …

Adaptive Resilient Control For A Class Of Nonlinear Distributed Parameter Systems With Actuator Faults, Hasan Ferdowsi, Jia Cai, Sarangapani Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

This paper presents a new model-based fault resilient control scheme for a class of nonlinear distributed parameter systems (DPS) represented by parabolic partial differential equations (PDE) in the presence of actuator faults. A Luenberger-like observer on the basis of nonlinear PDE representation of DPS is developed with boundary measurements. A detection residual is generated by taking the difference between the measured output of the DPS and the estimated one given by the observer. Once a fault is detected, an unknown actuator fault parameter vector together with a known basis function is utilized to adaptively estimate the fault dynamics. A novel …

Meta-Icvi: Ensemble Validity Metrics For Concise Labeling Of Correct, Under- Or Over-Partitioning In Streaming Clustering, Niklas M. Melton, Sasha Petrenko, Donald C. Wunsch

Electrical and Computer Engineering Faculty Research & Creative Works

Understanding the performance and validity of clustering algorithms is both challenging and crucial, particularly when clustering must be done online. Until recently, most validation methods have relied on batch calculation and have required considerable human expertise in their interpretation. Improving real-time performance and interpretability of cluster validation, therefore, continues to be an important theme in unsupervised learning. Building upon previous work on incremental cluster validity indices (iCVIs), this paper introduces the Meta- iCVI as a tool for explainable and concise labeling of partition quality in online clustering. Leveraging a time-series classifier and data-fusion techniques, the Meta- iCVI combines the outputs …

Malware Detection With Artificial Intelligence: A Systematic Literature Review, Matthew G. Gaber, Mohiuddin Ahmed, Helge Janicke

Research outputs 2022 to 2026

In this survey, we review the key developments in the field of malware detection using AI and analyze core challenges. We systematically survey state-of-the-art methods across five critical aspects of building an accurate and robust AI-powered malware-detection model: malware sophistication, analysis techniques, malware repositories, feature selection, and machine learning vs. deep learning. The effectiveness of an AI model is dependent on the quality of the features it is trained with. In turn, the quality and authenticity of these features is dependent on the quality of the dataset and the suitability of the analysis tool. Static analysis is fast but is …

A Systematic Review Of K-12 Cybersecurity Education Around The World, Ahmed Ibrahim, Marnie Mckee, Leslie F. Sikos, Nicola F. Johnson

Research outputs 2022 to 2026

This paper presents a systematic review of K-12 cybersecurity education literature from around the world. 24 academic papers dated from 2013-2023 were eligible for inclusion in the literature established within the research protocol. An additional 19 gray literature sources comprised the total. A range of recurring common topics deemed as aspects of cybersecurity behavior or practice were identified. A variety of cybersecurity competencies and skills are needed for K-12 students to apply their knowledge. As may be expected to be the case with interdisciplinary fields, studies are inherently unclear in the use of their terminology, and this is compounded in …

Empowering Causal Machine Learning For Large-Scale Manufacturing Pipelines With Knowledge Graphs, Yuxin Zi, Cory Henson, Amit P. Sheth

Faculty Publications

Understanding causal relations within manufacturing pipelines is crucial for key manufacturing tasks such as anomaly detection and root cause analysis. However, existing causal machine learning (causal ML) approaches struggle to scale effectively to the vast number of variables present in manufacturing settings. We advocate for incorporating domain knowledge within the manufacturing pipelines, represented as knowledge graphs (KGs), for designing causal ML methods for large-scale manufacturing problems. Knowledge graphs can encode rich contextual information about the interactions and dependencies between different components and stages of the manufacturing pipeline, providing a structured framework to guide the discovery of causal relationships. By incorporating …

Lifelong Learning-Based Optimal Trajectory Tracking Control Of Constrained Nonlinear Affine Systems Using Deep Neural Networks, Irfan Ganie, Sarangapani Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

This article presents a novel lifelong integral reinforcement learning (LIRL)-based optimal trajectory tracking scheme using the multilayer (MNN) or deep neural network (Deep NN) for the uncertain nonlinear continuous-time (CT) affine systems subject to state constraints. A critic MNN, which approximates the value function, and a second NN identifier are together used to generate the optimal control policies. The weights of the critic MNN are tuned online using a novel singular value decomposition (SVD)-based method, which can be extended to MNN with the N-hidden layers. Moreover, an online lifelong learning (LL) scheme is incorporated with the critic MNN to mitigate …

Online Continual Safe Reinforcement Learning-Based Optimal Control Of Mobile Robot Formations, Irfan Ganie, S. Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

In this work, a leader-follower tracking and formation control strategy for mobile robots (MRs) with uncertain dynamics is proposed. This strategy utilizes a continual lifelong safe reinforcement learning (CLSRL) framework based on multilayer neural networks (MNNs). The proposed design employs actor-critic MNNs, incorporating a barrier function. This function is derived from the Bellman optimality principle. It addresses the state constraints throughout the control design process. A novel online continual lifelong learning (CLL) method is introduced for MR formation. This method leverages the Bellman residual error for weight significance in MNNs. It addresses catastrophic forgetting and interlayer dependence through layer-specific regularizers. …

Reinforcement Learning-Based Constrained Optimal Control Of Strict-Feedback Nonlinear Systems: Application To Autonomous Underwater Vehicles, Behzad Farzanegan, S. Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

This paper addresses a constrained neural network (NN)-based optimal tracking scheme for a class of uncertain nonlinear discrete-time systems in strict-feedback form by using a control barrier function (CBF). First, a modified barrier-type cost function is introduced for each subsystem, guiding the actual system trajectory toward the safe set or desired trajectory while avoiding unwanted sets. To address the tracking problem, an augmented system is employed to convert the time-varying optimal tracking to a time-invariant optimal regulation. Then, an actor-critic framework is employed with the backstepping technique to obtain both virtual and actual optimal control policies for each subsystem to …

Learning From The Past: Using Peer Data To Improve Course Recommendations In Personalized Education, Colton Walker, Sahra Sedigh Sarvestani, Ali R. Hurson

Electrical and Computer Engineering Faculty Research & Creative Works

This research introduces a recommendation system designed to enhance student success by intelligently personalizing the semester schedules and graduation path based on the student's performance, interests, and background; and inspired by the academic journeys of similar students who have successfully graduated in the past. The proposed recommender system leverages a combination of Markov decision processes, Q-Learning, and collaborative filtering techniques to identify graduation paths with a higher likelihood of success for the student. The proposed model is versatile and generic and can be adapted to various disciplines if sufficient past historical data is available. The proposed model has been prototyped …

Survey Of Hidden Markov Models (Hmms) For Sign Language Recognition (Slr), Iwan Sandjaja, Ahmad Alsharoa, Donald Wunsch, Jian Liu

Electrical and Computer Engineering Faculty Research & Creative Works

This paper surveyed several significant papers on specific topics applying the Hidden Markov Models (HMMs) for Sign Language Recognition (SLR), divided into five main episodes: Classical HMMs, Extended HMMs, HMMs and Machine Learning, HMMs and Sensor Fusion, and HMMs and Big Data. This stringent survey would contribute significantly to advanced research on unification brain models such as neural networks, adaptive resonance theory, and confabulation theory. First, the HMM was introduced as one of the popular methods of performing SLR, and each episode of its development was expounded. In each episode, a main paper and several supporting papers were summarized. Next, …

Relative Altitude Estimation Of Infrared Thermal Uav Images Using Sift Features, Shirin Nasr Esfahani, Jagannathan Sarangapani

Electrical and Computer Engineering Faculty Research & Creative Works

Unmanned Aerial Vehicles (UAVs) have become indispensable in various applications, including surveillance, urban scene analysis, and agricultural monitoring. Accurate altitude estimation is critical for UAV operations, especially in environments where traditional sensors like GPS, pressure altimeters, and radar may fail. This paper explores the use of infrared and thermal imaging for relative altitude estimation of UAVs, highlighting their significant advantages over traditional RGB images. Infrared and thermal imaging offer superior performance in low-light and adverse weather conditions, providing clearer visibility and more reliable feature detection. By leveraging the Scale-Invariant Feature Transform (SIFT) features, this approach utilizes the inherent benefits of …

Lifelong Direct Error-Driven Learning For Uav Altitude Estimation In Different Weather Conditions, Shirin Nasr-Esfahani, Jagannathan Sarangapani

Electrical and Computer Engineering Faculty Research & Creative Works

While deep neural networks achieve remarkable visual perception capabilities for UAV position and orientation estimation, their resilience to different weather conditions still needs improvement. These models often suffer from catastrophic forgetting when adapted to new environments, losing previously acquired knowledge. Lifelong learning methods aim to balance learning flexibility and memory stability. In this paper, we present an image-based approach to estimate the relative altitude of a UAV using 2D images under varying weather conditions, including sunny, sunset, and foggy scenarios. Our experiments demonstrate significant performance degradation when the model is trained sequentially on different weather datasets, especially when new images …

Optimal Trajectory Tracking For Uncertain Linear Discrete-Time Systems Using Time-Varying Q-Learning, Maxwell Geiger, Vignesh Narayanan, Sarangapani Jagannathan

Electrical and Computer Engineering Faculty Research & Creative Works

This Article Introduces a Novel Optimal Trajectory Tracking Control Scheme Designed for Uncertain Linear Discrete-Time (DT) Systems. in Contrast to Traditional Tracking Control Methods, Our Approach Removes the Requirement for the Reference Trajectory to Align with the Generator Dynamics of an Autonomous Dynamical System. Moreover, It Does Not Demand the Complete Desired Trajectory to Be Known in Advance, Whether through the Generator Model or Any Other Means. Instead, Our Approach Can Dynamically Incorporate Segments (Finite Horizons) of Reference Trajectories and Autonomously Learn an Optimal Control Policy to Track Them in Real Time. to Achieve This, We Address the Tracking Problem …

Analysing Natural Language Processing Techniques: A Comparative Study Of Nltk, Spacy, Bert, And Distilbert On Customer Query Datasets., Patrizia De Camillis

ICT

This study examines the role of sentiment analysis in customer queries, emphasising its impact on brand perception and the risks of poor query management. It compares the performance of NLP models—NLTK, spaCy, BERT, and DistilBERT—on customer query and feedback data. The findings show that BERT and DistilBERT produce similar results, often categorising queries as neutral, indicating their strength in handling diverse sentiments. NLTK and spaCy also share performance patterns. The research offers insights into the capabilities and limitations of these models in sentiment analysis.