Theory and Algorithms Commons^™

Dynamic-Query Robustness Of Ann Indexes Under Time-Indexed Drift, Stellamaris Nakacwa, Majid Shaalan

Harrisburg University Other Works

Approximate nearest-neighbor search is a central retrieval primitive in dense question-answering and retrieval-augmented generation systems. Existing ANN evaluation protocols typically measure recall, latency, throughput, and search-effort sensitivity under a fixed-query assumption: a query vector is submitted to an index, approximate neighbors are retrieved, and the result is compared with exact nearest-neighbour ground truth. This assumption is appropriate for conventional vector-search benchmarking, but it is less complete for multi-step, distributed, and agent-controlled retrieval pipelines in which the retrieval-facing query may be refined, recomputed, or displaced across execution steps. This paper introduces a time-driven dynamic query evaluation framework for ANN search. The …

Online Visual Query System For Real-Time Large-Scale Spatio-Temporal Data Explorations With Error Bounds, Xueqi Huang

Dissertations, Theses, and Capstone Projects

Modern datasets continue to grow in size, dimensionality, and heterogeneity, creating increasing tension between the need for responsive, interactive analysis and the computational cost of accessing, aggregating, and visualizing large volumes of data. Traditional database engines and visualization tools often assume that full data retrieval is feasible or that exact computation is necessary for meaningful insight. In practice, however, analysts frequently benefit from timely, uncertainty-aware approximations than from delayed and exact results. This thesis investigates how data summarization techniques, specifically mergeable sketches can be combined with progressive, out-of-core visualization methods to support interactive exploration of datasets that exceed main memory. …

Toward Efficient And Scalable Scientific Data Management Through Quality-Oriented Data Compression, Pu Jiao

Theses and Dissertations--Computer Science

Scientific simulations and instruments now produce data at rates that overwhelm the storage, memory, and network subsystems of modern high-performance computing (HPC) facilities. Error-bounded lossy compression reduces data movement costs while bounding reconstruction error, yet three barriers limit its adoption in mission-critical workflows: existing compressors cannot guarantee the accuracy of domain-specific quantities of interest (QoIs) derived from compressed data; compression-induced artifacts such as posterization, blocking, and interpolation banding erode user confidence in decompressed fields; and significant compressibility in the quantization index arrays of interpolation-based pipelines remains unexploited. This dissertation addresses all three barriers through four contributions, with the artifact barrier …

Universal Systems Simulation Via Constraint Hypergraphs With Applications To Digital Twins, John Morris

All Dissertations

The characterization of systems encompasses a variety of modeling frameworks designed to capture specific behaviors and components of various system domains. Whatever the framework, the core elements of a system representation are the information of the system and a description of how that information is related. The relations in deterministic systems are functions, which, when composed to form executable processes, can be used to simulate system data. A declarative modeling framework is one that encodes mechanisms for preparing these simulations within the model structure, allowing an external agent to form the execution processes required for a given context. To date, …

Topic Modeling And Culturomic Analysis Of 30,000 Books Over 100 Years Using Gensim, Michael A. Freeman

Electronic Theses and Dissertations

This thesis explores the cultural influence of historical events on English-language fiction published between 1820 and 1929. Using a corpus of 30,256 digitized books from Project Gutenberg, Latent Dirichlet Allocation (LDA) topic modeling was applied to identify recurring themes across eleven decades. The study sought to determine whether historically significant events could be detected within fictional narratives. One clear instance emerged: Napoleon Bonaparte and the Napoleonic Wars appeared explicitly in the 1820s corpus. Beyond this, several thematic patterns were observed—such as maritime language in the 1840s, national identity in the 1880s, and youth-oriented dialogue in the early 20th century—that plausibly …

Island-Based Evolutionary Computation With Diverse Surrogates And Adaptive Knowledge Transfer For High-Dimensional Data-Driven Optimization, Xianrong Zhang, Yuejiao Gong, Zhiguang Cao, Jun Zhang

Research Collection School Of Computing and Information Systems

In recent years, there has been a growing interest in data-driven evolutionary algorithms (DDEAs) employing surrogate models to approximate the objective functions with limited data. However, current DDEAs are primarily designed for lower-dimensional problems and their performance drops significantly when applied to large-scale optimization problems (LSOPs). To address the challenge, this paper proposes an offline DDEA named DSKT-DDEA. DSKT-DDEA leverages multiple islands that utilize different data to establish diverse surrogate models, fostering diverse subpopulations and mitigating the risk of premature convergence. In the intra-island optimization phase, a semi-supervised learning method is devised to fine-tune the surrogates. It not only facilitates …

Non-Homophilic Graph Pre-Training And Prompt Learning, Xingtong Yu, Jie Zhang, Yuan Fang, Renhe Jiang

Research Collection School Of Computing and Information Systems

Graphs are ubiquitous for modeling complex relationships between objects across various fields. Graph neural networks (GNNs) have become a mainstream technique for graph-based applications, but their performance heavily relies on abundant labeled data. To reduce labeling requirement, pre-training and prompt learning has become a popular alternative. However, most existing prompt methods do not distinguish between homophilic and heterophilic characteristics in graphs. In particular, many real-world graphs are non-homophilic-neither strictly nor uniformly homophilic-as they exhibit varying homophilic and heterophilic patterns across graphs and nodes. In this paper, we propose ProNoG, a novel pre-training and prompt learning framework for such non-homophilic graphs. …

Syntax-Enhanced Boundary-Aware Named Entity Recognition Model, Chuanming Yu, Bin Deng, Zhengang Zhang

Journal of Scientific Information Research

[Purpose/significance] This study addresses the issue of inadequate perception of entity boundaries in traditional character-level modeling-based named entity recognition models by integrating syntax information containing entity boundary features into the task using a multi-head graph attention network with dense connections. This integration enhances the effectiveness of named entity recognition.

[Method/process] This study proposes a Syntax-enhanced Boundary-aware Named Entity Recognition Model (SynBNER), which utilizes BERT for text semantic representation and integrates syntax information using a dense-connected graph attention network. This integration incorporates implicit entity boundary information from syntax information into word representations, thereby enhancing the model's entity boundary perception capability.

[Result/conclusion] …

Adapting Large Language Models For Parameter-Efficient Log Anomaly Detection, Ying Fu Lim, Jiawen Zhu, Guansong Pang

Research Collection School Of Computing and Information Systems

Log Anomaly Detection (LAD) seeks to identify atypical patterns in log data that are crucial to assessing the security and condition of systems. Although Large Language Models (LLMs) have shown tremendous success in various fields, the use of LLMs in enabling the detection of log anomalies is largely unexplored. This work aims to fill this gap. Due to the prohibitive costs involved in fully fine-tuning LLMs,we explore the use of parameter-efficient fine-tuning techniques (PEFTs) for adapting LLMs to LAD.To have an in-depth exploration of the potential of LLM-driven LAD, we present a comprehensive investigation of leveraging two of the most …

Less Is More: On The Importance Of Data Quality For Unit Test Generation, Junwei Zhang, Xing Hu, Shan Gao, Xin Xia, David Lo, Shanping Li

Research Collection School Of Computing and Information Systems

Unit testing is crucial for software development and maintenance. Effective unit testing ensures and improves software quality, but writing unit tests is time-consuming and labor-intensive. Recent studies have proposed deep learning (DL) techniques or large language models (LLMs) to automate unit test generation. These models are usually trained or fine-tuned on large-scale datasets. Despite growing awareness of the importance of data quality, there has been limited research on the quality of datasets used for test generation. To bridge this gap, we systematically examine the impact of noise on the performance of learning-based test generation models. We first apply the open …

Unraveling Genetic Links Between Diabetes And Heart Failure-A Machine Learning Approach, Sunakhi Sahoo, Marzieh Ayati

Research Symposium

Background: Diabetic heart failure (DHF) is defined as a chronic and progressive disease which is associated with both diabetes and heart failure (HF). Even though there have been many developments in the knowledge of these diseases, there is still much to learn about the genetic crossovers between the two. In this study, we identified genes that are associated with diabetic heart failure and heart failure by using gene expression data from patients with DHF, HF, and a control group of patients who died of natural causes. We sought to identify genes that had altered expression levels which could possibly play …

Bloom: Behavioral Learning And Outcome Observation In Microbes, Sean Sarwar Haque, Luke Compton Wharton, Ming Lin, Razvan Voicu

Symposium of Student Scholars

Understanding how pathogens respond to physical changes in their environment is crucial for developing effective treatments and preventative measures. Current research often relies on static models or experimental data that either fail to capture the dynamic interactions within cellular environments or are not generalizable to other types of pathogens. This project aims to address this gap by creating a comprehensive cell simulation that models pathogens and their response to chemical, physical, and physiological changes. The proposed solution is a simulation that integrates biological data and computational modeling to replicate the behavior of pathogens in real time as they are affected …

Towards Smart Farming: Image-Based Crop Health Assessment And Disease Diagnosis Using Deep Learning Techniques, Kristina Botova

Master's Theses or Doctor of Nursing Practice

Accurate crop monitoring is essential for optimizing agricultural productivity and ensuring food security. This study presents a comprehensive deep learning framework for image crop type recognition, health status prediction, and disease detection using multiple Convolutional Neural Network (CNN) models. The proposed approach uses open-source datasets consisting of five crop types (apple, corn, grape, potato, tomato), varying health conditions, and common diseases. By deploying specialized CNN architecture focused on each task, the system achieves a high accuracy of 99.25% in classifying crop types, identifying health status, and detecting specific diseases. Compared to a single CNN model, the use of the proposed …

Visualization Of Paleocurrents On A Web Application Using Gplates, Anjan Sapkota

MS in Computer Science Theses

Paleocurrents are flow directions derived from features of sedimentary rocks that reveal the direction of the current of wind or water that deposited the sediment. In 2015, Brand et al. created a global database of paleocurrents, which contains over 1,000,000 measurements worldwide: North America, South America, Australia, Great Britain, parts of Western Europe, China, Africa are fairly well represented; Antarctica, Eastern Europe, and Asia are modestly represented and Russia is poorly represented. The contribution of this thesis is a web application that uses the GPlates’ Application Programming Interface (API) to visualize global paleocurrents through time in an interactive way based …

Harnessing Collective Structure Knowledge In Data Augmentation For Graph Neural Networks, Rongrong Ma, Guansong Pang, Ling Chen

Research Collection School Of Computing and Information Systems

Graph neural networks (GNNs) have achieved state-of-the-art performance in graph representation learning. Message passing neural networks, which learn representations through recursively aggregating information from each node and its neighbors, are among the most commonly-used GNNs. However, a wealth of structural information of individual nodes and full graphs is often ignored in such process, which restricts the expressive power of GNNs. Various graph data augmentation methods that enable the message passing with richer structure knowledge have been introduced as one main way to tackle this issue, but they are often focused on individual structure features and difficult to scale up with …

Unraveling The Dynamics Of Stable And Curious Audiences In Web Systems, Rodrigo Alves, Antoine Ledent, Renato Assunção, Pedro Vaz-De-Melo, Marius Kloft

Research Collection School Of Computing and Information Systems

We propose the Burst-Induced Poisson Process (BPoP), a model designed to analyze time series data such as feeds or search queries. BPoP can distinguish between the slowly-varying regular activity of a stable audience and the bursty activity of a curious audience, often seen in viral threads. Our model consists of two hidden, interacting processes: a self-feeding process (SFP) that generates bursty behavior related to viral threads, and a non-homogeneous Poisson process (NHPP) with step function intensity that is influenced by the bursts from the SFP. The NHPP models the normal background behavior, driven solely by the overall popularity of the …

Ft2ra: A Fine-Tuning-Inspired Approach To Retrieval-Augmented Code Completion, Qi Guo, Shangqing Liu, Xiaofei Xie, Ze Tang Tang

Research Collection School Of Computing and Information Systems

The rise of code pre-trained models has significantly enhanced various coding tasks, such as code completion, and tools like GitHub Copilot. However, the substantial size of these models, especially large models, poses a significant challenge when it comes to fine-tuning them for specific downstream tasks. As an alternative approach, retrieval-based methods have emerged as a promising solution, augmenting model predictions without the need for fine-tuning. Despite their potential, a significant challenge is that the designs of these methods often rely on heuristics, leaving critical questions about what information should be stored or retrieved and how to interpolate such information for …

Robust Asynchronous Federated Learning With Time-Weighted And Stale Model Aggregation, Yinbin Miao, Ziteng Liu, Xinghua Li, Meng Li, Hongwei Li, Kim-Kwang Raymond Choo, Robert H. Deng

Research Collection School Of Computing and Information Systems

Federated Learning (FL) ensures collaborative learning among multiple clients while maintaining data locally. However, the traditional synchronous FL solutions have lower accuracy and require more communication time in scenarios where most devices drop out during learning. Therefore, we propose an Asynchronous Federated Learning (AsyFL) scheme using time-weighted and stale model aggregation, which effectively solves the problem of poor model performance due to the heterogeneity of devices. Then, we integrate Symmetric Homomorphic Encryption (SHE) into AsyFL to propose Asynchronous Privacy-Preserving Federated Learning (Asy-PPFL), which protects the privacy of clients and achieves lightweight computing. Privacy analysis shows that Asy-PPFL is indistinguishable under …

Prompt Tuning On Graph-Augmented Low-Resource Text Classification, Zhihao Wen, Yuan Fang

Research Collection School Of Computing and Information Systems

Text classification is a fundamental problem in information retrieval with many real-world applications, such as predicting the topics of online articles and the categories of e-commerce product descriptions. However, low-resource text classification, with no or few labeled samples, presents a serious concern for supervised learning. Meanwhile, many text data are inherently grounded on a network structure, such as a hyperlink/citation network for online articles, and a user-item purchase network for e-commerce products. These graph structures capture rich semantic relationships, which can potentially augment low-resource text classification. In this paper, we propose a novel model called Graph-Grounded Pre-training and Prompting (G2P2) …

Machine Learning: Face Recognition, Mohammed E. Amin

Publications and Research

This project explores the cutting-edge intersection of machine learning (ML) and face recognition (FR) technology, utilizing the OpenCV library to pioneer innovative applications in real-time security and user interface enhancement. By processing live video feeds, our system encodes visual inputs and employs advanced face recognition algorithms to accurately identify individuals from a database of photos. This integration of machine learning with OpenCV not only showcases the potential for bolstering security systems but also enriches user experiences across various technological platforms. Through a meticulous examination of unique facial features and the application of sophisticated ML algorithms and neural networks, our project …

Comparative Predictive Analysis Of Stock Performance In The Tech Sector, Asaad Sendi

LSU New Orleans Theses and Dissertations

This study compares the performance of deep learning models, including Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Transformer, in predicting stock prices across five companies (AAPL, CSCO, META, MSFT, and TSLA) from July 2019 to July 2023. Key findings reveal that GRU models generally exhibit the lowest Mean Absolute Error (MAE), indicating higher precision, particularly notable for CSCO with a remarkably low MAE. While LSTM models often show slightly higher MAE values, they outperform Transformer models in capturing broader trends and variance in stock prices, as evidenced by higher R-squared (R2) values. Transformer models generally exhibit higher MAE …

Knowledge Enhanced Multi-Intent Transformer Network For Recommendation, Ding Zou, Wei Wei, Feida Zhu, Chuanyu Xu, Tao Zhang, Chengfu Huo

Research Collection School Of Computing and Information Systems

Incorporating Knowledge Graphs (KGs) into Recommendation has attracted growing attention in industry, due to the great potential of KG in providing abundant supplementary information and interpretability for the underlying models. However, simply integrating KG into recommendation usually brings in negative feedback in industry, mainly due to the ignorance of the following two factors: i) users' multiple intents, which involve diverse nodes in KG. For example, in e-commerce scenarios, users may exhibit preferences for specific styles, brands, or colors. ii) knowledge noise, which is a prevalent issue in Knowledge Enhanced Recommendation (KGR) and even more severe in industry scenarios. The irrelevant …

Quantum Machine Learning For Credit Scoring, Nikolaos Schetakis, Davit Aghamalyan, Micheael Boguslavsky, Agnieszka Rees, Marc Rakotomalala, Paul Robert Griffin

Research Collection School Of Computing and Information Systems

This study investigates the integration of quantum circuits with classical neural networks for enhancing credit scoring for small- and medium-sized enterprises (SMEs). We introduce a hybrid quantum–classical model, focusing on the synergy between quantum and classical rather than comparing the performance of separate quantum and classical models. Our model incorporates a quantum layer into a traditional neural network, achieving notable reductions in training time. We apply this innovative framework to a binary classification task with a proprietary real-world classical credit default dataset for SMEs in Singapore. The results indicate that our hybrid model achieves efficient training, requiring significantly fewer epochs …

Adaptive Content-Aware Influence Maximization Via Online Learning To Rank, Konstantinos Theocharidis, Panagiotis Karras, Manolis Terrovitis, Spiros Skiadopoulos, Hady Wirawan Lauw

Research Collection School Of Computing and Information Systems

How can we adapt the composition of a post over a series of rounds to make it more appealing in a social network? Techniques that progressively learn how to make a fixed post more influential over rounds have been studied in the context of the Influence Maximization (IM) problem, which seeks a set of seed users that maximize a post’s influence. However, there is no work on progressively learning how a post’s features affect its influence. In this article, we propose and study the problem of Adaptive Content-Aware Influence Maximization (ACAIM), which calls to find k features to form a …

Multi-Modality Transformer For E-Commerce: Inferring User Purchase Intention To Bridge The Query-Product Gap, Srivatsa Mallapragada

Dissertations

The rapid growth of e-commerce has necessitated the development of sophisticated product retrieval systems that can effectively match user queries with relevant products. However, the semantic gap between queries and products remains a significant challenge, as traditional retrieval methods often fail to capture the nuances of user purchase intentions. E-commerce click-stream data and product catalogs offer critical user behavior insights and product knowledge that are untapped in the current product search algorithms. This dissertation presents learning strategies that leverage the query-product transaction logs to enrich the pipeline of our proposed multi-modal transformer model, which transforms initial user queries into pseudo …

Faster Rates For Compressed Federated Learning With Client-Variance Reduction, Haoyu Zhao, Konstantin Burlachenko, Zhize Li, Peter Richtarik

Research Collection School Of Computing and Information Systems

Due to the communication bottleneck in distributed and federated learning applications, algorithms using communication compression have attracted significant attention and are widely used in practice. Moreover, the huge number, high heterogeneity, and limited availability of clients result in high client -variance. This paper addresses these two issues together by proposing compressed and clientvariance reduced methods COFIG and FRECON. We prove an O( (1+\omega)3/2\surdN+ (1+\omega)N2/3 S\epsilon2 S\epsilon2 ) bound on the number of communication rounds of COFIG in the nonconvex setting, where N is the total number of clients, S is the number of clients participating in each round, \epsilon is …

Screening Through A Broad Pool: Towards Better Diversity For Lexically Constrained Text Generation, Changsen Yuan, Heyan Huang, Yixin Cao, Qianwen Cao

Research Collection School Of Computing and Information Systems

Lexically constrained text generation (CTG) is to generate text that contains given constrained keywords. However, the text diversity of existing models is still unsatisfactory. In this paper, we propose a lightweight dynamic refinement strategy that aims at increasing the randomness of inference to improve generation richness and diversity while maintaining a high level of fluidity and integrity. Our basic idea is to enlarge the number and length of candidate sentences in each iteration, and choose the best for subsequent refinement. On the one hand, different from previous works, which carefully insert one token between two words per action, we insert …

Music Genre Classification Capabilities Of Enhanced Neural Network Architectures, Joshua Engelkes

Scholarly Horizons: University of Minnesota, Morris Undergraduate Journal

With the increase of digital music audio uploads, applications that deal with music information have been widely requested by streaming platforms. Automatic music genre classification is an important function of music recommendation and music search applications. Since the music genre categorization criteria continually shift, data-driven methods such as neural networks have been proven especially useful to music information retrieval. An enhanced CNN architecture, the Bottom-up Broadcast Neural Network, uses mel-spectrograms to push music data through a network where important low-level information is preserved. An enhanced RNN architecture, the Independent Recurrent Neural Network for Music Genre Classification, takes advantage of the …

Imitate The Good And Avoid The Bad: An Incremental Approach To Safe Reinforcement Learning, Minh Huy Hoang, Mai Anh Tien, Pradeep Varakantham

Research Collection School Of Computing and Information Systems

A popular framework for enforcing safe actions in Reinforcement Learning (RL) is Constrained RL, where trajectory based constraints on expected cost (or other cost measures) are employed to enforce safety and more importantly these constraints are enforced while maximizing expected reward. Most recent approaches for solving Constrained RL convert the trajectory based cost constraint into a surrogate problem that can be solved using minor modifications to RL methods. A key drawback with such approaches is an over or underestimation of the cost constraint at each state. Therefore, we provide an approach that does not modify the trajectory based cost constraint …

Recommendations With Minimum Exposure Guarantees: A Post-Processing Framework, Ramon Lopes, Rodrigo Alves, Antoine Ledent, Rodrygo L. T. Santos, Marius Kloft

Research Collection School Of Computing and Information Systems

Relevance-based ranking is a popular ingredient in recommenders, but it frequently struggles to meet fairness criteria because social and cultural norms may favor some item groups over others. For instance, some items might receive lower ratings due to some sort of bias (e.g. gender bias). A fair ranking should balance the exposure of items from advantaged and disadvantaged groups. To this end, we propose a novel post-processing framework to produce fair, exposure-aware recommendations. Our approach is based on an integer linear programming model maximizing the expected utility while satisfying a minimum exposure constraint. The model has fewer variables than previous …