Digital Commons Network^™

No Experts, No Problem: Avoidance Learning From Bad Demonstrations, Minh Huy Hoang, Tien Mai, Pradeep Varakantham

Research Collection School Of Computing and Information Systems

This paper addresses the problem of learning avoidance behavior within the context of offline imitation learning. In contrast to conventional methodologies that prioritize the replication of expert or near-expert demonstrations, our work investigates a setting where expert (or desirable) data is absent, and the objective is to learn to eschew undesirable actions by leveraging demonstrations of such behavior (i.e., learning from negative examples).To address this challenge, we propose a novel training objective grounded in the maximum entropy principle. We further characterize the fundamental properties of this objective function, reformulating the learning process as a cooperative inverse Q-learning task. Moreover, we …

Bias Testing And Mitigation In Llm-Based Code Generation, Dong Huang, Jie M. Zhang, Qingwen Bu, Xiaofei Xie, Junjie Chen, Heming Cui

Research Collection School Of Computing and Information Systems

As the adoption of LLMs becomes more widespread in software coding ecosystems, a pressing issue has emerged: does the generated code contain social bias and unfairness, such as those related to age, gender, and race? This issue concerns the integrity, fairness, and ethical foundation of software applications that depend on the code generated by these models but are underexplored in the literature. This paper presents a novel bias testing framework that is specifically designed for code generation tasks. Based on this framework, we conduct an extensive empirical study on the biases in code generated by five widely studied LLMs (i.e., …

Kpiroot+: An Efficient Integrated Framework For Anomaly Detection And Root Cause Analysis In Large-Scale Cloud Systems, Wenwei Gu, Renyi Zhong, Guangba Yu, Xinying Sun, Jinyang Liu, Yintong Huo, Zhuangbin Chen, Jianping Zhang, Jiazhen Gu, Yongqiang Yang, Michael R. Lyu

Research Collection School Of Computing and Information Systems

To ensure the reliability of cloud systems, their runtime status reflecting the service quality is periodically monitored with monitoring metrics, i.e., KPIs (key performance indicators). When performance issues happen, root cause localization pinpoints the specific KPIs that are responsible for the degradation of overall service quality, facilitating prompt problem diagnosis and resolution. To this end, existing methods generally locate root-cause KPIs by identifying the KPIs that exhibit a similar anomalous trend to the overall service performance. While straightforward, solely relying on the similarity calculation may be ineffective when dealing with cloud systems with complicated interdependent services. Recent deep learning-based methods …

Genscore: Agent-Based Short-Answer Question Generation And Scoring In Software Engineering Courses, Nguyen Binh Duong Ta, Lwin Khin Shar

Research Collection School Of Computing and Information Systems

Short-answer questions are commonly used in educational assessments, as they are often viewed as a more effective way than multiple-choice questions to determine whether students have achieved the intended learning outcomes. However, manually creating appropriate questions targeting different cognitive levels such as those defined by the Bloom’s Taxonomy, and grading text answers from students are not trivial tasks for instructors. Existing work on auto-question generation and scoring in computing education typically targets coding-based questions. However, in software engineering courses, assessments can extend beyond coding to understanding of processes, DevOps methodologies, system design, etc. This work aims to address the dual …

Large Language Models (Llms) For Clinical Note Generation: International Classification Of Disease (Icd) Code, Knowledge Graph (Kg) And Prompt Evaluation, Ivan P. Makohon

Computer Science Theses & Dissertations

In the past decade, a surge in the amount of electronic health record (EHR) data in the United States occurred, driven by a favorable policy environment created by the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 and the 21st Century Cures Act of 2016. Clinical notes for patients’ assessments, diagnoses, and treatments are captured in these EHRs in free-form text by physicians, who spend a considerable amount of time entering them. Manually writing these notes is time-consuming, increasing patient waiting times and potentially delaying diagnoses. Large language models (LLMs), such as GPT-4o, possess the ability …

Copyright Ownership And Duration Of Ai-Authored Works, Cheng Lim Saw

Research Collection Yong Pung How School Of Law

On the assumption that Parliament has endorsed the notion of AI authorship and the prospect that copyright may well subsist in works created autonomously by the AI itself, this essay further explores allied issues surrounding the ownership and duration of copyright in AI-authored works.

Hybrid-Balance Gflownet For Solving Vehicle Routing Problems, Ni Zhang, Zhiguang Cao

Research Collection School Of Computing and Information Systems

Existing GFlowNet-based methods for vehicle routing problems (VRPs) typically employ Trajectory Balance (TB) to achieve global optimization but often neglect important aspects of local optimization. While Detailed Balance (DB) addresses local optimization more effectively, it alone falls short in solving VRPs, which inherently require holistic trajectory optimization. To address these limitations, we introduce the Hybrid-Balance GFlowNet (HBG) framework, which uniquely integrates TB and DB in a principled and adaptive manner by aligning their intrinsically complementary strengths. Additionally, we propose a specialized inference strategy for depot-centric scenarios like the Capacitated Vehicle Routing Problem (CVRP), leveraging the depot node's greater flexibility in …

Multi-Task Vehicle Routing Solver Via Mixture Of Specialized Experts Under State-Decomposable Mdp, Yuxin Pan, Zhiguang Cao, Chengyang Gu, Liu Liu, Peilin Zhao, Yize Chen, Fangzhen Lin

Research Collection School Of Computing and Information Systems

Existing neural methods for multi-task vehicle routing problems (VRPs) typically learn unified solvers to handle multiple constraints simultaneously. However, they often underutilize the compositional structure of VRP variants, each derivable from a common set of basis VRP variants. This critical oversight causes unified solvers to miss out the potential benefits of basis solvers, each specialized for a basis VRP variant. To overcome this limitation, we propose a framework that enables unified solvers to perceive the shared-component nature across VRP variants by proactively reusing basis solvers, while mitigating the exponential growth of trained neural solvers. Specifically, we introduce a State-Decomposable MDP …

Mando-Llm: Heterogeneous Graph Transformers With Large Language Models For Smart Contract Vulnerability Detection, Nhat Minh Nguyen, Huu Hoang Nguyen, Long Le Thanh, Zahra Ahmadi, Thanh Nam Doan, Daoyuan Wu, Lingxiao Jiang

Research Collection School Of Computing and Information Systems

Detecting vulnerabilities in smart contracts is vital for the security and reliability of decentralized apps. To facilitate vulnerability detection, contract codes, including bug patterns, are represented as heterogeneous graphs with various nodes and edges, like control-flow and function-call graphs. However, existing graph learning techniques struggle with large, complex graphs. This paper presents MANDO-LLM, a novel framework that combines heterogeneous graph transformers (HGTs) with large language models (LLMs) for detecting vulnerabilities in smart contracts represented as heterogeneous contract graphs built upon control-flow and call graphs. MANDO-LLM uses LLMs to capture code features from control-flow and call data, customizes HGTs to learn …

Stableguard: Towards Unified Copyright Protection And Tamper Localization In Latent Diffusion Models, Haoxin Yang, Bangzhen Liu, Xuemiao Xu, Cheng Xu, Yuyang Yu, Zikai Huang, Yi Wang, Shengfeng He

Research Collection School Of Computing and Information Systems

The advancement of diffusion models has enhanced the realism of AI-generated content but also raised concerns about misuse, necessitating robust copyright protection and tampering localization. Although recent methods have made progress toward unified solutions, their reliance on post hoc processing introduces considerable application inconvenience and compromises forensic reliability. We propose StableGuard, a novel framework that seamlessly integrates a binary watermark into the diffusion generation process, ensuring copyright protection and tampering localization in Latent Diffusion Models through an end-to-end design. We develop a Multiplexing Watermark VAE (MPW-VAE) by equipping a pretrained Variational Autoencoder (VAE) with a lightweight latent residual-based adapter, enabling …

Rising From Ashes: Generalized Federated Learning Via Dynamic Parameter Reset, Jiahao Wu, Ming Hu, Yanxin Yang, Xiaofei Xie, Zekai Chen, Chenyu Song, Mingsong Chen

Research Collection School Of Computing and Information Systems

Although Federated Learning (FL) is promising for privacy-preserving collaborative model training, it suffers from low inference performance due to heterogeneous client data. Due to heterogeneous data across clients, FL training easily learns client-specific overfitting features. Existing FL methods adopt coarsegrained averaging, which can easily cause the global model to get stuck in local optima, leading to poor generalization. Specifically, this paper presents a novel FL framework, FedPhoenix, to address this issue. It stochastically resets partial parameters in each round to destroy some features of the global model, guiding FL training to learn multiple generalized features for inference rather than specific …

When Less Language Is More: Language-Reasoning Disentanglement Makes Llms Better Multilingual Reasoners, Weixiang Zhao, Jiahe Guo, Yang Deng, Tongtong Wu, Wenxuan Zhang, Yulin Hu, Xingyu Sui, Yanyan Zhao, Wanxiang Che, Bing Qin, Tat-Seng Chua, Ting Liu

Research Collection School Of Computing and Information Systems

Multilingual reasoning remains a significant challenge for large language models (LLMs), with performance disproportionately favoring high-resource languages. Drawing inspiration from cognitive neuroscience, which suggests that human reasoning functions largely independently of language processing, we hypothesize that LLMs similarly encode reasoning and language as separable components that can be disentangled to enhance multilingual reasoning. To evaluate this, we perform a causal intervention by ablating language-specific representations at inference time. Experiments on 10 open-weight LLMs spanning 11 typologically diverse languages show that this language-specific ablation consistently boosts multilingual reasoning performance. Layer-wise analyses further confirm that language and reasoning representations can be effectively …

Efskip: A New Error Feedback With Linear Speedup For Compressed Federated Learning With Arbitrary Data Heterogeneity, Hongyan Bao, Pengwen Chen, Ying Sun, Zhize Li

Research Collection School Of Computing and Information Systems

Due to the communication bottleneck in distributed and decentralized federated learning applications, algorithms using compressed communication have attracted significant attention. The Error Feedback (EF) is a widely-studied compression framework for convergence with biased compressors such as top-k sparsification. Although various improvements have been obtained in recent years, the theoretical guarantee for EF-type framework is still limited. Previous works either 1) rely on strong assumptions such as bounded gradient/dissimilarity assumptions, thus can not deal with arbitrary data heterogeneity and also slow the convergence speed, or 2) can not enjoy linear speedup in the number of clients. In this work, we propose …

Generalization Bounds For Rank‑Sparse Neural Networks, Antoine Ledent, Rodrigo Alves, Yunwen Lei

Research Collection School Of Computing and Information Systems

It has been recently observed in much of the literature that neural networks exhibit a bottleneck rank property: for larger depths, the activation and weights of neural networks trained with gradient-based methods tend to be of approximately low rank. In fact, the rank of the activations of each layer converges to a fixed value referred to as the “bottleneck rank”, which is the minimum rank required to represent the training data. This perspective is in line with the observation that regularizing linear networks (without activations) with weight decay is equivalent to minimizing the Schatten p quasi norm of the neural …

Sciteuq: Toward Uncertainty-Aware Complex Scientific Table Data Extraction And Understanding, Kehinde Ajayi

Computer Science Theses & Dissertations

Scientific tables report critical research insights, data, and findings for scientific progress. Because Portable Document Format (PDF) is the de facto standard format for scientific paper publishing, there has been an emerging need for an automatic method to extract data from PDF files. A significant fraction of scientific tables exhibit complex structure and content, making it challenging for machine learning tools to accurately extract the content directly from PDF files. Despite the advancements in Table Structure Recognition (TSR), automated extraction of data from complex scientific tables remains a challenge due to variations in table structures and contents. In this dissertation, …

Making Explanations Make Sense: Xai For Smishing Detection, Eleni Alexandra Katsarakes

Psychology Theses & Dissertations

Explainable Artificial Intelligence (XAI) is a key component of effective human-AI collaboration, particularly in high-stakes domains such as cybersecurity. While AI tools hold promise for mitigating threats such as SMS-based phishing (SMiShing), their real-world effectiveness may hinge not just on detection accuracy, but on whether users can make sense of the system’s outputs. As SMiShing attacks grow in both frequency and sophistication, so does the urgency of designing human-centered AI systems that support user decision-making under uncertainty. This study examined how four distinct AI explanation types - Normative (rule-based), Attributive (feature-based), Exemplar (case-based), and Recommendation-Only - influence user performance, confidence, …

Human Identification And Action Recognition Using Small Data And Deep Domain Adaptation, Alexander M. Glandon

Electrical & Computer Engineering Theses & Dissertations

Human identification and human action recognition problems are two important research areas for real-world security and surveillance applications. In both human identification and action recognition, it is necessary to operate by collecting small datasets in the field, possibly in a short time window of observation. This dissertation studies and develops computational modeling and high-performance machine learning (ML) and deep learning (DL) models for human identification and human action recognition using small amounts of data. These methods and computational models may be useful for different security and surveillance applications.

This dissertation on human recognition develops a ML computational model to estimate …

Toward Personalizing Quantum Computing Education: An Evolutionary Llm-Powered Approach, Iizalaarab Elhaimeur

Computer Science Theses & Dissertations

Quantum computing education faces significant challenges due to its complexity and the limitations of current tools. This thesis introduces a novel Intelligent Teaching Assistant for quantum computing education and details its evolutionary design process. The system combines a knowledge-graph-augmented architecture with two specialized LLM agents: a Teaching Agent for dynamic interaction and a Lesson Planning Agent for lesson generation. The system is designed to adapt to individual student needs, with interactions meticulously tracked and stored in a knowledge graph. This graph represents student actions, learning resources, and their relationships, aiming to enable reasoning about effective learning pathways. We describe the …

Early Conceptual Sketches Of Blended Reality And The Precursor To The Bbs Quad (2022), David Smith

Publications and Research

This document contains two original hand-drawn conceptual sketches created in early 2022, representing the earliest visual formulations of what would later evolve into the Balanced Blended Space (BBS) framework. The drawings predate my first conversations with ChatGPT and were produced as part of my independent sabbatical research into blended environments, mediated performance, and human–machine interaction.

The first drawing examines human–computational mediation, perception, and internal mapping. The second sketch—later referred to informally as the “BBS Quad”—extends this idea by reconciling cognition–computation symmetry with physical–virtual spatial relationships. Published together, these images document the conceptual foundations of the BBS framework prior to its …

Multi-Agent Robotaxi Dispatch Coordination In A Real-World Simulation – Optimizing Rider Assignment, Rebalancing, And Charging Using Battery-Dependent Rewards And Welfare Maximization, Paden Thompson

All Graduate Theses and Dissertations, Fall 2023 to Present

We propose an approach to coordinate a robotaxi fleet for an autonomous ride-hail service. This is a service similar to a traditional ride-hailing service (Uber, Lyft), where customers request a ride and are then picked up in a car and dropped off in a new location; except, driverless vehicles called robotaxis are used to transport the customers.

Our approach teaches helpful coordination strategies to a robotaxi fleet while taking into account the individual battery level of the robotaxis. Each robotaxi acts as an individual agent in our simulation and can choose to pick up a rider, reposition to a new …

Cross-Modal Prompting For Multi-Class Visual Anomaly Localization, Duncan F. Mccain

All Theses

Visual anomaly detection is a technology that uses computer vision to automatically identify defects or irregularities in images, such as cracks, scratches, or discolorations on manufactured products. Unsupervised visual anomaly detection does this without needing examples of those defects during the training process. This "unsupervised" approach is crucial in industries like manufacturing, automotive, electronics, and pharmaceuticals, where ensuring product quality is essential for safety, reliability, and cost efficiency. For instance, it helps spot flaws in circuit boards, fabrics, or medical pills during production lines, preventing faulty items from reaching consumers. By reducing manual inspections, it saves time and resources, benefiting …

From Digital Divide To Equity-Enhancing Diffusion: Generative Ai And Writing Quality, Rebecca Tukachinsky Forster, Kerk Kee, Gabriel Miao Li

Communication Faculty Articles and Research

This study investigates whether generative AI can narrow the gap between stronger and developing writers and explores the mechanisms underlying these effects. In a within-subject experiment, students wrote two essays, with and without AI assistance. Computer-aided analysis of the writing quality confirmed that while all students benefited from AI, that less skillful writers gained more. There was also no evidence of skillful writers using AI in more sophisticated and beneficial ways. The study contributes to theorizing the digital divide and offers insights into maximizing the benefits of AI tools. Theoretically, we situate generative-AI use within Diffusion of Innovations, treating ChatGPT …

Benchmarking Dna Foundation Models For Genomic And Genetic Tasks, Haonan Feng, Lang Wu, Bingxin Zhao, Chad Huff, Jianjun Zhang, Jia Wu, Lifeng Lin, Peng Wei, Chong Wu

School of Medicine Faculty Publications

The rapid evolution of DNA foundation models promises to revolutionize genomics, yet comprehensive evaluations are lacking. Here, we present a comprehensive, unbiased benchmark of five models (DNABERT-2, Nucleotide Transformer V2, HyenaDNA, Caduceus-Ph, and GROVER) across diverse genomic and genetic tasks including sequence classification, gene expression prediction, variant effect quantification, and topologically associating domain (TAD) region recognition, using zero-shot embeddings. Our analysis reveals that mean token embedding consistently and significantly improves sequence classification performance, outperforming other pooling strategies. Model performance varies among tasks and datasets; while general purpose DNA foundation models showed competitive performance in pathogenic variant identification, they were less …

Ai-Based Mapping Of Offshore Wind Energy Around The Korean Peninsula Using Sentinel-1 Sar And Numerical Weather Prediction Data, Jason Sung-Uk Joh, Son V. Nghiem, Menas Kafatos, Jay Liu, Jinsoo Kim, Seung Hee Kim

Institute for ECHO Articles and Research

Offshore wind farm projects are being promoted in the seas surrounding the Korean Peninsula to secure renewable energy. To support site selection, offshore wind resource maps were generated using deep neural networks trained on Sentinel-1 SAR imagery, numerical weather prediction data, offshore wind observations, sea surface temperature, and bathymetry. The deep neural network (DNN) framework consisted of six sub-models targeting eastward and northward wind components across three regions—the Yellow Sea, Korea Strait, and East Sea—to account for spatial heterogeneity. The proposed models outperformed existing approaches, achieving mean absolute errors (MAE) ranging from 1.31 to 1.69 m/s and correlation coefficients (CC) …

Introduction To Machine Learning And Machine Learning Systems, Raffi T. Khatchadourian Ph.D.

Open Educational Resources

Lecture slides introducing machine learning and machine learning systems for an undergraduate software engineering course. Topics include what machine learning is and how it differs from traditional programming, foundation models, the major types of learning (supervised, unsupervised, reinforcement, and others), and applications across domains. Using a food-delivery time-prediction case study, the deck walks through a typical ML pipeline—data collection and cleaning, feature engineering, model training, and evaluation—and covers evaluation methods (precision and recall, confusion matrices, error measures) along with underfitting versus overfitting and the realities of learning and evaluation in production. Based on "Machine Learning in Production/AI Engineering" by Christian …

Ai-Driven Optimization Of Wind Energy Distribution In Texas Using Multi-Agent Reinforcement Learning, Waleed Amer, Owolabi Oluwadamilola, Bassey Ogbonnaya

SMU Data Science Review

Abstract. The integration of large-scale wind power into modern electrical grids presents persistent challenges due to variability, curtailment, and compliance with operational constraints. This study proposes a multi-agent reinforcement learning (MARL) framework for optimizing wind energy distribution within the Texas power grid. The system employs three specialized agents—managing wind curtailment, storage utilization, and load adjustments—to collaboratively balance supply and demand under dynamic grid conditions. Using historical operational data from the Electric Reliability Council of Texas (ERCOT), the framework was trained and evaluated on a range of scenarios encompassing both typical and extreme operating conditions. Results demonstrate substantial performance improvements compared …

Archaeological Predictive Meta-Modeling In Pre-Columbian Mexico, Peter Stamm

Electronic Theses and Dissertations

Archaeological Predictive Modeling stands firmly as an important tool for Archaeologists to predict undiscovered sites from civilizations all across the globe. While powerful, this methodology is not without its own set of qualms. Striking a balance between pure a data-driven approach while also observing leading expert theories can be a complicated task. Going further, deciding on the specific domain of features to emphasize or overlook can be a challenge within itself, as one misstep can drastically change the output of model, sometimes for the worst. In addition, creating models that can expose their reasoning process can be rather difficult to …

Unless Ai Washes The Dishes, Can We Really Call It Intelligent?, Essraa Nawar

Library Articles and Research

"A few weeks ago, I found myself sitting with a question that keeps resurfacing as AI becomes louder, faster, and everywhere. What happens when the models know everything about our lives except the one thing that matters most in the moment. It is remarkable how much of human decision making is driven not by external information but by internal states. A tightening in the chest. A sudden clarity. A quiet discomfort that redirects us before we can explain why. These signals guide our choices in ways computation cannot replicate."

Fair Use In The Age Of Generative Ai: Navigating Copyright Challenges In Educational Contexts, Wendy Wallberg

Faculty and Staff Publications & Presentations

Generative AI tools are everywhere, but what’s actually allowed when it comes to copyright and teaching? This session breaks down what fair use means in the age of AI, covers current legal cases, and offers practical tools to help educators and institutions use AI responsibly and confidently.

A Usv Path Planning Algorithm Under Special Environment Based On Td3-Rrt, Jitong Chen, Jiajia Zhou, Di Wu, Hailong Jiang

Journal of System Simulation

Abstract: In view of USV path planning in special environments such as multiple obstacles, large-size obstacles, and narrow passages, the rapidly-exploring random tree (RRT) algorithm suffers from drawbacks such as a large sampling base, low success rate, and zigzagging planned path. To address these problems, a global path planning algorithm (TD3-RRT) was proposed based on the twin delayed deep deterministic policy gradient (TD3). The USV path search model was established by combining the RRT algorithm with deep reinforcement learning. Forward looking detection was used to sense the environment to adaptively adjust the step size. The path search direction was exported …