Old Dominion University
Southern Methodist University
Bowling Green State University
The College of Wooster
University of Pennsylvania
University of Massachusetts Amherst
James Madison University
San Jose State University
Learning Adversarial Semantic Embeddings For Zero-Shot Recognition In Open Worlds,
2024
Singapore Management University
Learning Adversarial Semantic Embeddings For Zero-Shot Recognition In Open Worlds, Tianqi Li, Guansong Pang, Xiao Bai, Jin Zheng, Lei Zhou, Xin Ning
Research Collection School Of Computing and Information Systems
Zero-Shot Learning (ZSL) focuses on classifying samples of unseen classes with only their side semantic information presented during training. It cannot handle real-life, open-world scenarios where there are test samples of unknown classes for which neither samples (e.g., images) nor their side semantic information is known during training. Open-Set Recognition (OSR) is dedicated to addressing the unknown class issue, but existing OSR methods are not designed to model the semantic information of the unseen classes. To tackle this combined ZSL and OSR problem, we consider the case of “Zero-Shot Open-Set Recognition” (ZS-OSR), where a model is trained under the ZSL …
Diffusion-Based Negative Sampling On Graphs For Link Prediction,
2024
Singapore Management University
Diffusion-Based Negative Sampling On Graphs For Link Prediction, Yuan Fang, Yuan Fang
Research Collection School Of Computing and Information Systems
Link prediction is a fundamental task for graph analysis with important applications on the Web, such as social network analysis and recommendation systems, etc. Modern graph link prediction methods often employ a contrastive approach to learn robust node representations, where negative sampling is pivotal. Typical negative sampling methods aim to retrieve hard examples based on either predefined heuristics or automatic adversarial approaches, which might be inflexible or difficult to control. Furthermore, in the context of link prediction, most previous methods sample negative nodes from existing substructures of the graph, missing out on potentially more optimal samples in the latent space. …
On The Feasibility Of Simple Transformer For Dynamic Graph Modeling,
2024
Singapore Management University
On The Feasibility Of Simple Transformer For Dynamic Graph Modeling, Yuxia Wu, Yuan Fang, Lizi Liao
Research Collection School Of Computing and Information Systems
Dynamic graph modeling is crucial for understanding complex structures in web graphs, spanning applications in social networks, recommender systems, and more. Most existing methods primarily emphasize structural dependencies and their temporal changes. However, these approaches often overlook detailed temporal aspects or struggle with long-term dependencies. Furthermore, many solutions overly complicate the process by emphasizing intricate module designs to capture dynamic evolutions. In this work, we harness the strength of the Transformer’s self-attention mechanism, known for adeptly handling long-range dependencies in sequence modeling. Our approach offers a simple Transformer model, called SimpleDyG, tailored for dynamic graph modeling without complex modifications. We …
Optimization Of Memory Management Using Machine Learning,
2024
Southern Adventist University
Optimization Of Memory Management Using Machine Learning, Luke Bartholomew
Campus Research Day
This paper is a proposed solution to the problem of memory safety using machine learning. Memory overload and corruption cause undesirable behaviors in a system that are addressed by memory safety implementations. This project uses machine learning models to classify different states of system memory from a dataset collected from a Raspberry Pi System. These models can then be used to classify real run time memory data and increase memory safety overall in a system.
Innovating Inventory And Alert Systems With Object Tracking,
2024
Southern Adventist University
Innovating Inventory And Alert Systems With Object Tracking, Juan Harmse, Esther Peden
Campus Research Day
Security system users require safeguarding inventory from potential theft while reducing manual tracking of physical objects. Our contribution harnesses the power of artificial intelligence and computer vision with YOLO to automate the process of tracking inventory items. The system sends alerts to the inventory manager when it detects particular events. Our approach was evaluated with KernProf profiling, interference, and orientation tests. The results were overall positive in these testing areas.
The Vulnerabilities Of Artificial Intelligence Models And Potential Defenses,
2024
William & Mary
The Vulnerabilities Of Artificial Intelligence Models And Potential Defenses, Felix Iov
Cybersecurity Undergraduate Research Showcase
The rapid integration of artificial intelligence (AI) into various commercial products has raised concerns about the security risks posed by adversarial attacks. These attacks manipulate input data to disrupt the functioning of AI models, potentially leading to severe consequences such as self-driving car crashes, financial losses, or data breaches. We will explore neural networks, their weaknesses, and potential defenses. We will discuss adversarial attacks including data poisoning, backdoor attacks, evasion attacks, and prompt injection. Then, we will explore defense strategies such as data protection, input sanitization, and adversarial training. By understanding how adversarial attacks work and the defenses against them, …
Comprehensive Question And Answer Generation With Llama 2,
2024
Southern Adventist University
Comprehensive Question And Answer Generation With Llama 2, Matous Hybl
MS in Computer Science Theses
Since the introduction of transformers, large language models have proven capable in many natural language processing fields. However, existing systems still face challenges in generating high-quality extractive questions. Base models and public chatbots fall short if the question source or quantity are critical. Our contribution is a question and answer generator for generating comprehensive, extractive questions and answers. This approach includes fine-tuning a LLaMA 2 base model for answer extraction (AE) and question generation (QG). We evaluate the resulting system using common automated metrics and a manual evaluation. We find that our system is comparable to the latest research and …
Immersive Japanese Language Learning Web Application Using Spaced Repetition, Active Recall, And An Artificial Intelligent Conversational Chat Agent Both In Voice And In Text,
2024
Southern Adventist University
Immersive Japanese Language Learning Web Application Using Spaced Repetition, Active Recall, And An Artificial Intelligent Conversational Chat Agent Both In Voice And In Text, Marc Butler
MS in Computer Science Project Reports
In the last two decades various human language learning applications, spaced repetition software, online dictionaries, and artificial intelligent chat agents have been developed. However, there is no solution to cohesively combine these technologies into a comprehensive language learning application including skills such as speaking, typing, listening, and reading. Our contribution is to provide an immersive language learning web application to the end user which combines spaced repetition, a study technique used to review information at systematic intervals, and active recall, the process of purposely retrieving information from memory during a review session, with an artificial intelligent conversational chat agent both …
High-Resolution And Quality Settings With Latent Consistency Models,
2024
Old Dominion University
High-Resolution And Quality Settings With Latent Consistency Models, Steven Chen, Junrui Zhang, Rui Ning
Cybersecurity Undergraduate Research Showcase
Diffusion Models have become powerful generative models which is capable of synthesizing high-quality images across various domains. This paper explores Stable Diffusion and mostly focuses on Latent Diffusion Models. Latent Consistency Models can enhance the inference with minimal iterations. It demonstrates the performance in image in-painting and class-conditional synthesis tasks. Throughout the experiment different datasets and parameter configurations, the paper highlights the image quality, processing time, and parameter. It also discussed the future directions including adding trigger-based implementation and emotional-based themes to replace the prompt.
The Security Of Deep Neural Networks,
2024
Norfolk State University
The Security Of Deep Neural Networks, Jalaya Allen
Cybersecurity Undergraduate Research Showcase
Our society has transitioned from our primitive lifestyle to soon, an increasingly automatic one. That idea is further exemplified as we shift into an AI era, better known as Artificial intelligence. Artificial Intelligence is classified as computer systems that can perform tasks that typically require human intelligence. However, a common thought or question that most might have is, how is this done? How does AI process information the way we want it to and have access to so much information? AI is trained by systems called AI models. These modeling programs are trained on data to recognize patterns or make …
Artificial Sociality,
2024
University of Turin, Italy
Artificial Sociality, Simone Natale, Iliana Depounti
Human-Machine Communication
This article proposes the notion of Artificial Sociality to describe communicative AI technologies that create the impression of social behavior. Existing tools that activate Artificial Sociality include, among others, Large Language Models (LLMs) such as ChatGPT, voice assistants, virtual influencers, socialbots and companion chatbots such as Replika. The article highlights three key issues that are likely to shape present and future debates about these technologies, as well as design practices and regulation efforts: the modelling of human sociality that foregrounds it, the problem of deception and the issue of control from the part of the users. Ethical, social and cultural …
Artificial Intelligence Could Probably Write This Essay Better Than Me,
2024
Augustana College, Rock Island Illinois
Artificial Intelligence Could Probably Write This Essay Better Than Me, Claire Martino
Augustana Center for the Study of Ethics Essay Contest
No abstract provided.
Navigating The Maze: The Role Of Pre-Enrollment Socio-Cultural And Institutional Factors In Higher Education In The Age Of Ai,
2024
Lindenwood University
Navigating The Maze: The Role Of Pre-Enrollment Socio-Cultural And Institutional Factors In Higher Education In The Age Of Ai, Emily Barnes, James Hutson
Faculty Scholarship
This article explores the complex interplay between pre-enrollment socio-cultural and institutional factors and their impact on the higher education landscape. It challenges traditional metrics of academic achievement, presenting a nuanced perspective on student success that emphasizes the importance of socio-economic backgrounds, cultural capital, and K-12 education quality. The analysis extends to the significant role of institutional attributes in shaping student readiness and decision-making processes. The study advocates for the integration of artificial intelligence (AI)-driven assessments by higher education institutions to cater to the diverse needs of the student body, promoting an inclusive and supportive learning environment. Anchored in an extensive …
Multi-Aspect Rule-Based Ai: Methods, Taxonomy, Challenges And Directions Towards Automation, Intelligence And Transparent Cybersecurity Modeling For Critical Infrastructures,
2024
Edith Cowan University
Multi-Aspect Rule-Based Ai: Methods, Taxonomy, Challenges And Directions Towards Automation, Intelligence And Transparent Cybersecurity Modeling For Critical Infrastructures, Iqbal H. Sarker, Helge Janicke, Mohamed A. Ferrag, Alsharif Abuadbba
Research outputs 2022 to 2026
Critical infrastructure (CI) typically refers to the essential physical and virtual systems, assets, and services that are vital for the functioning and well-being of a society, economy, or nation. However, the rapid proliferation and dynamism of today's cyber threats in digital environments may disrupt CI functionalities, which would have a debilitating impact on public safety, economic stability, and national security. This has led to much interest in effective cybersecurity solutions regarding automation and intelligent decision-making, where AI-based modeling is potentially significant. In this paper, we take into account “Rule-based AI” rather than other black-box solutions since model transparency, i.e., human …
Combating Financial Crimes With Unsupervised Learning Techniques: Clustering And Dimensionality Reduction For Anti-Money Laundering,
2024
Faculty of Science Al-Azhar University Cairo, Egypt
Combating Financial Crimes With Unsupervised Learning Techniques: Clustering And Dimensionality Reduction For Anti-Money Laundering, Ahmed N. Bakry, Almohammady S. Alsharkawy, Mohamed S. Farag, Kamal R. Raslan
Al-Azhar Bulletin of Science
Anti-Money Laundering (AML) is a crucial task in ensuring the integrity of financial systems. One keychallenge in AML is identifying high-risk groups based on their behavior. Unsupervised learning, particularly clustering, is a promising solution for this task. However, the use of hundreds of features todescribe behavior results in a highdimensional dataset that negatively impacts clustering performance.In this paper, we investigate the effectiveness of combining clustering method agglomerative hierarchicalclustering with four dimensionality reduction techniques -Independent Component Analysis (ICA), andKernel Principal Component Analysis (KPCA), Singular Value Decomposition (SVD), Locality Preserving Projections (LPP)- to overcome the issue of high-dimensionality in AML data and …
Graph Neural Network Guided By Feature Selection And Centrality Measures For Node Classification On Homophilic And Heterophily Graphs,
2024
Department of Mathematics, Faculty of Science, Al-Azhar University, Cairo, Egypt.
Graph Neural Network Guided By Feature Selection And Centrality Measures For Node Classification On Homophilic And Heterophily Graphs, Asmaa M. Mahmoud, Heba F. Eid, Abeer S. Desuky, Hoda A. Ali
Al-Azhar Bulletin of Science
One of the most recent developments in the fields of deep learning and machine learning is Graph Neural Networks (GNNs). GNNs core task is the feature aggregation stage, which is carried out over the node's neighbours without taking into account whether the features are relevant or not. Additionally, the majority of these existing node representation techniques only consider the network's topology structure while completely ignoring the centrality information. In this paper, a new technique for explaining graph features depending on four different feature selection approaches and centrality measures in order to identify the important nodes and relevant node features is …
Environmental, Social, And Governance (Esg) And Artificial Intelligence In Finance: State-Of-The-Art And Research Takeaways,
2024
Singapore Management University
Environmental, Social, And Governance (Esg) And Artificial Intelligence In Finance: State-Of-The-Art And Research Takeaways, Trstan Lim
Research Collection School Of Computing and Information Systems
The rapidly growing research landscape in finance, encompassing environmental, social, and governance (ESG) topics and associated Artificial Intelligence (AI) applications, presents challenges for both new researchers and seasoned practitioners. This study aims to systematically map the research area, identify knowledge gaps, and examine potential research areas for researchers and practitioners. The investigation focuses on three primary research questions: the main research themes concerning ESG and AI in finance, the evolution of research intensity and interest in these areas, and the application and evolution of AI techniques specifically in research studies within the ESG and AI in finance domain. Eight archetypical …
Exploring The Potential Of Chatgpt In Automated Code Refinement: An Empirical Study,
2024
Singapore Management University
Exploring The Potential Of Chatgpt In Automated Code Refinement: An Empirical Study, Qi Guo, Shangqing Liu, Junming Cao, Xiaohong Li, Xin Peng, Xiaofei Xie, Bihuan Chen
Research Collection School Of Computing and Information Systems
Code review is an essential activity for ensuring the quality and maintainability of software projects. However, it is a time-consuming and often error-prone task that can significantly impact the development process. Recently, ChatGPT, a cutting-edge language model, has demonstrated impressive performance in various natural language processing tasks, suggesting its potential to automate code review processes. However, it is still unclear how well ChatGPT performs in code review tasks. To fill this gap, in this paper, we conduct the first empirical study to understand the capabilities of ChatGPT in code review tasks, specifically focusing on automated code refinement based on given …
Extracting Dnn Architectures Via Runtime Profiling On Mobile Gpus,
2024
University of Massachusetts Amherst
Extracting Dnn Architectures Via Runtime Profiling On Mobile Gpus, Dong Hyub Kim
Masters Theses
Due to significant investment, research, and development efforts over the past decade, deep neural networks (DNNs) have achieved notable advancements in classification and regression domains. As a result, DNNs are considered valuable intellectual property for artificial intelligence providers. Prior work has demonstrated highly effective model extraction attacks which steal a DNN, dismantling the provider’s business model and paving the way for unethical or malicious activities, such as misuse of personal data, safety risks in critical systems, or spreading misinformation. This thesis explores the feasibility of model extraction attacks on mobile devices using aggregated runtime profiles as a side-channel to leak …
Policy Gradient Methods: Analysis, Misconceptions, And Improvements,
2024
University of Massachusetts Amherst
Policy Gradient Methods: Analysis, Misconceptions, And Improvements, Christopher P. Nota
Doctoral Dissertations
Policy gradient methods are a class of reinforcement learning algorithms that optimize a parametric policy by maximizing an objective function that directly measures the performance of the policy. Despite being used in many high-profile applications of reinforcement learning, the conventional use of policy gradient methods in practice deviates from existing theory. This thesis presents a comprehensive mathematical analysis of policy gradient methods, uncovering misconceptions and suggesting novel solutions to improve their performance. We first demonstrate that the update rule used by most policy gradient methods does not correspond to the gradient of any objective function due to the way the …
