Digital Commons Network^™

Demo2test: Transfer Testing Of Agent In Competitive Environment With Failure Demonstrations, Jianming Chen, Yawen Wang, Junjie Wang, Xiaofei Xie, Dandan Wang, Qing Wang, Fanjiang Xu

Research Collection School Of Computing and Information Systems

The competitive game between agents exists in many critical applications, such as military unmanned aerial vehicles. It is urgent to test these agents to reduce the significant losses caused by their failures. Existing studies mainly are to construct a testing agent that competes with the target agent to induce its failures. These approaches usually focus on a single task, requiring much more time for multi-task testing. However, if the previously tested tasks (source tasks) and the task to be tested (target task) share similar agents or task objectives, the transferable knowledge in source tasks can potentially increase the effectiveness of …

Neuron Semantic-Guided Test Generation For Deep Neural Networks Fuzzing, Li Huang, Weifeng Sun, Meng Yan, Zhongxin Liu, Yan Lei, David Lo

Research Collection School Of Computing and Information Systems

In recent years, significant progress has been made in testing methods for deep neural networks (DNNs) to ensure their correctness and robustness. Coverage-guided criteria, such as neuron-wise, layer-wise, and path-/trace-wise, have been proposed for DNN fuzzing. However, existing coverage-based criteria encounter performance bottlenecks for several reasons: Testing Adequacy: Partial neural coverage criteria have been observed to achieve full coverage using only a small number of test inputs. In this case, increasing the number of test inputs does not consistently improve the quality of models. Interpretability: The current coverage criteria lack interpretability. Consequently, testers are unable to identify and understand which …

Interpreting Topic Models In Byte-Pair Encoding Space, Jia Peng Lim, Hady Wirawan Lauw

Research Collection School Of Computing and Information Systems

Byte-pair encoding (BPE) is pivotal for processing text into chunksize tokens, particularly in Large Language Model (LLM). From a topic modeling perspective, as these chunksize tokens might be mere parts of valid words, evaluating and interpreting these tokens for coherence is challenging. Most, if not all, of coherence evaluation measures are incompatible as they benchmark using valid words. We propose to interpret the recovery of valid words from these tokens as a ranking problem and present a model-agnostic and training-free recovery approach from the topic-token distribution onto a selected vocabulary space, following which we could apply existing evaluation measures. Results …

Weakly-Supervised Semantic Segmentation With Image-Level Labels: From Traditional Models To Foundation Models, Zhaozheng Chen, Qianru Sun

Research Collection School Of Computing and Information Systems

The rapid development of deep learning has driven significant progress in image semantic segmentation—a fundamental task in computer vision. Semantic segmentation algorithms often depend on the availability of pixel-level labels (i.e., masks of objects), which are expensive, time consuming, and labor intensive. Weakly supervised semantic segmentation (WSSS) is an effective solution to avoid such labeling. It utilizes only partial or incomplete annotations and provides a cost-effective alternative to fully supervised semantic segmentation. In this article, our focus is on the WSSS with image-level labels, which is the most challenging form of WSSS. Our work has two parts. First, we conduct …

Synthesizing Multi-Person And Rare Pose Images For Human Pose Estimation, Liuqing Zhao, Zichen Tian, Zou Peng, Richang Hong, Qianru Sun

Research Collection School Of Computing and Information Systems

Human pose estimation (HPE) models underperform in recognizing rare poses because they suffer from data imbalance problems (i.e., there are few image samples for rare poses) in their training datasets. From a data perspective, the most intuitive solution is to synthesize data for rare poses. Specifically, the rule-based methods apply manual manipulations (such as Cutout and GridMask) to the existing data, so the limited diversity of the data constrains the model. An alternative method is to learn the underlying data distribution via deep generative models (such as ControlNet and HumanSD) and then sample “new data” from the distribution. This works …

Dims: Distributed Index For Similarity Search In Metric Spaces, Yifan Zhu, Chengyang Luo, Tang Qian, Lu Chen, Yunjun Gao, Baihua Zheng

Research Collection School Of Computing and Information Systems

Similarity search finds objects that are similar to a given query object based on a similarity metric. As the amount and variety of data continue to grow, similarity search in metric spaces has gained significant attention. Metric spaces can accommodate any type of data and support flexible distanc e metrics, making similarity search in metric spaces beneficial for many real-world applications, such as multimedia retrieval, personalized recommendation, trajectory analytics, data mining, decision planning, and distributed servers. However, existing studies mostly focus on indexing metric spaces on a single machine, which faces efficiency and scalability limitations with increasing data volume and …

Transforming Urban Dynamics: Harnessing Large Language Models For Smarter Mobility, Hao Xue, Ming Jin, Shirui Pan, Flora Salim, Guansong Pang

Research Collection School Of Computing and Information Systems

Artificial intelligence (AI) has the potential to analyze mobility data and make mobility systems smarter by leveraging diverse data sources such as geospatial data, transportation logs, and real-time sensor data to optimize traffic flow, enhance public transportation systems, and support the development of autonomous vehicles. With the newly emerged generative AI paradigm, exemplified by large language models (LLMs), there is great potential to transform the current AI applications in mobility, transportation, and urban domains. This article provides an overview of recent efforts and aims to shed light on the challenges and future opportunities to facilitate the adaptation of LLMs for …

Attribute-Based Encryption With Searchable Encryption, Yang Yang

Research Collection School Of Computing and Information Systems

Attribute-based encryption (ABE) with searchable encryption is a notion that inherits the advantages of both ABE (Goyal et al., 2006) and searchable encryption (SE) (Boneh et al., 2004; Boneh and Waters, 2007) mechanisms to realize secure share and search for the outsourced data. ABE creates fine-grained access control system to prevent sensitive data from being accessed by unauthorized user or external attacker (Bethencourt et al., 2007). SE generates encrypted index for outsourced data such that it can be searched by a keyword trapdoor (or token) (Shi et al., 2007; Naveed et al., 2014), where the outsourced content and search keyword(s) …

Adapting Installation Instructions In Rapidly Evolving Software Ecosystems, Haoyu Gao, Christoph Treude, Mansooreh Zahedi

Research Collection School Of Computing and Information Systems

files play an important role in providing installation-related instructions to software users and are widely used in open source software systems on platforms such as GitHub. Software projects evolve rapidly alongside their dependencies in dynamic software ecosystems, requiring frequent updates to installation instructions. These instructions are crucial for users to start with a software project. Despite their significance, there is a lack of systematic understanding regarding the documentation efforts invested in README files and the triggers behind them. To fill the research gap, we conducted a qualitative study, investigating 400 GitHub repositories with 1,163 README commits that focused on updates …

Adversarial Generative Flow Network For Solving Vehicle Routing Problems, Ni Zhang, Jingfeng Yang, Zhiguang Cao, Xu Chi

Research Collection School Of Computing and Information Systems

Recent research into solving vehicle routing problems (VRPs) has gained significant traction, particularly through the application of deep (reinforcement) learning for end-to-end solution construction. However, many current construction-based neural solvers predominantly utilize Transformer architectures, which can face scalability challenges and struggle to produce diverse solutions. To address these limitations, we introduce a novel framework beyond Transformer-based approaches, i.e., Adversarial Generative Flow Networks (AGFN). This framework integrates the generative flow network (GFlowNet)-a probabilistic model inherently adept at generating diverse solutions (routes)-with a complementary model for discriminating (or evaluating) the solutions. These models are trained alternately in an adversarial manner to improve …

Gnnsynergy: A Multi-View Graph Neural Network For Predicting Anti-Cancer Drug Synergy, Zhifeng Hao, Jianming Zhan, Yuan Fang, Min Wu, Ruichu Cai

Research Collection School Of Computing and Information Systems

Drug combinations play very important roles in cancer therapy, as they can enhance curative efficacy and overcome drug resistance. Due to the increasing size of combinatorial space, experimental screening for all the drug combinations becomes infeasible in practice. Therefore, there is a great need to develop accurate computational approaches that can predict potential drug combinations to direct the experimental screening. In this paper, we propose a novel method called GNNSynergy to learn drug embeddings for drug synergy prediction. Given a specific cancer cell line, we propose a multi-view graph neural network framework which considers the current cell line as main …

Double Oracle Neural Architecture Search For Game Theoretic Deep Learning Models, Aye Phyu Phyu Aung, Xinrun Wang, Ruiyu Wang, Hau Chan, Bo An, Xiaoli Li, J. Senthilnath

Research Collection School Of Computing and Information Systems

In this paper, we propose a new approach to train deep learning models using game theory concepts including Generative Adversarial Networks (GANs) and Adversarial Training (AT) where we deploy a double-oracle framework using best response oracles. GAN is essentially a two-player zero-sum game between the generator and the discriminator. The same concept can be applied to AT with attacker and classifier as players. Training these models is challenging as a pure Nash equilibrium may not exist and even finding the mixed Nash equilibrium is difficult as training algorithms for both GAN and AT have a large-scale strategy space. Extending our …

Recdreamer: Consistent Text-To-3d Generation Via Uniform Score Distillation, Chenxi Zheng, Yihong Lin, Bangzhen Liu, Xuemiao Xu, Yongwei Nie, Shengfeng He

Research Collection School Of Computing and Information Systems

Current text-to-3D generation methods based on score distillation often suffer from geometric inconsistencies, leading to repeated patterns across different poses of 3D assets. This issue, known as the Multi-Face Janus problem, arises because existing methods struggle to maintain consistency across varying poses and are biased toward a canonical pose. While recent work has improved pose control and approximation, these efforts are still limited by this inherent bias, which skews the guidance during generation. To address this, we propose a solution called RecDreamer, which reshapes the underlying data distribution to achieve more consistent pose representation. The core idea behind our method …

Label Correlated Contrastive Learning For Medical Report Generation, Xinyao Liu, Junchang Xin, Bing Tian Dai, Qi Shen, Zhihong Huang, Zhiqiong Wang

Research Collection School Of Computing and Information Systems

Background and Objective: Automatic generation of medical reports reduces both the burden on radiologists and the possibility of errors due to the inexperience of radiologists. The model that utilizes attention mechanism and contrastive learning can generate medical reports by capturing both general and specific semantics. However, existing contrastive learning methods ignore the specificity of medical data, that is, a patient may suffer from multiple diseases at the same time. This means that the lack of fine-grained relationships for contrastive learning will lead to the problem of insufficient specificity. Methods: To address the above problem, a label correlated contrastive learning method …

The Effectiveness Of Local Updates For Decentralized Learning Under Data Heterogeneity, Tongle Wu, Zhize Li, Ying Sun

Research Collection School Of Computing and Information Systems

We revisit two fundamental decentralized optimization methods, Decentralized Gradient Tracking (DGT) and Decentralized Gradient Descent (DGD), with multiple local updates. We consider two settings and demonstrate that incorporating local update steps can reduce communication complexity. Specifically, for  $\mu$-strongly convex and $L$-smooth loss functions, we proved that local DGT  achieves communication complexity {}{$\tilde{\mathcal{O}} \Big(\frac{L}{\mu(K+1)} + \frac{\delta + {}{\mu}}{\mu (1 - \rho)} + \frac{\rho }{(1 - \rho)^2} \cdot \frac{L+ \delta}{\mu}\Big)$}, where $K$ is the number of additional local update}, $\rho$ measures the network connectivity and $\delta$ measures the second-order heterogeneity of the local losses. Our results reveal the tradeoff between communication and …

Risk Spillover Effect Of China-Asean Supply Chains: Insights Of Industrial Transfer, Zeyang Bian, Yuning Zhang, Keng Siau, Yaqian Zhang, Jianjia He

Research Collection School Of Computing and Information Systems

As labour costs in China increase, labour-intensive industries are migrating to ASEAN countries, attracted by lower labour costs and market potential. This shift not only affects the economies of China and ASEAN but also reshapes the global manufacturing landscape. This paper investigates the correlation and spillover of supply chain risks using production exposure indicators derived from inter-country input-output data and the R-Vine Copula model. We assess the risk spillover of each country within the global supply chain. Our findings indicate that industrial relocation can significantly alter supply chain structures, thereby affecting the concentration and direction of risks. While China's role …

More Effective Javascript Breaking Change Detection Via Dynamic Object Relation Graph, Dezhen Kong, Jiakun Liu, Chao Ni, David Lo, Lingfeng Bao

Research Collection School Of Computing and Information Systems

JavaScript libraries are characterized by their widespread use, frequent code changes, and a high tolerance for backward incompatible changes. Awareness of such breaking changes can help developers adapt to version updates and avoid negative impacts. Several tools have been targeted to or can be used to detect breaking change detection in the JavaScript community. However, these tools detect breaking changes using different ways, and there are currently no systematic reviews of these approaches. From a preliminary study on popular JavaScript libraries, we find that existing approaches, including simple regression testing, model-based testing and type differencing cannot detect many breaking changes …

Charge Your Clients: Payable Secure Computation And Its Applications, Cong Zhang, Liqiang Peng, Weiran Liu, Shuaishuai Li, Meng Hao, Lei Zhang, Dongdai Lin

Research Collection School Of Computing and Information Systems

The online realm has witnessed a surge in the buying and selling of data, prompting the emergence of dedicated data marketplaces. These platforms cater to servers (sellers), enabling them to set prices for access to their data, and clients (buyers), who can subsequently purchase these data, thereby streamlining and facilitating such transactions. However, the current data market is primarily confronted with the following issues. Firstly, they fail to protect client privacy, presupposing that clients submit their queries in plaintext. Secondly, these models are susceptible to being impacted by malicious client behavior, for example, enabling clients to potentially engage in arbitrage …

Financial Named Entity Recognition: How Far Can Llm Go?, Yi-Te Lu, Yintong Huo

Research Collection School Of Computing and Information Systems

The surge of large language models (LLMs) has revolutionized the extraction and analysis of crucial information from a growing volume of financial statements, announcements, and business news. Recognition for named entities to construct structured data poses a significant challenge in analyzing financial documents and is a foundational task for intelligent financial analytics. However, how effective are these generic LLMs and their performance under various prompts are yet need a better understanding. To fill in the blank, we present a systematic evaluation of state-of-the-art LLMs and prompting methods in the financial Named Entity Recognition (NER) problem. Specifically, our experimental results highlight …

Habits And Preferences Of Filipino Viewers Of Local Films In Metro Manila, Ma. Elena C. Javier, Cristina A. Rodriguez, Jerome V. Cleofas, Zaldy C. Collado, John Joshua V. Barbo, Melvin A. Jabar, Jose Javier Reyes

SDRC Projects

The Philippine film industry continues to face challenges associated with declining cinema attendance, changing audience preferences, and the growing popularity of digital streaming platforms. These developments, intensified by the COVID-19 pandemic, have raised concerns about the sustainability of local film production and the ability of Filipino films to attract and retain audiences.

This research report examines the habits and preferences of film consumers in Metro Manila, focusing on the demographic, socioeconomic, cultural, and economic factors influencing cinema attendance and the patronage of Filipino films. Using a mixed-methods approach, the study found that 67% of surveyed moviegoers preferred streaming services, while …

Assessment On The Capacity Of Local Government Units In Implementing The Rules And Regulations Of Republic Act No. 11930 And A Cost Study Of Implementing An Anti-Osaec Program, Ma. Divina Gracia Roldan, Ador R. Torneo, Maria Caridad Tarroja, Marites M. Tiongco, Ma. Ella C. Oplas

SDRC Projects

This study assessed the capacity of selected local government units (LGUs) to implement Republic Act No. 11930, or the Anti-Online Sexual Abuse or Exploitation of Children and Anti-Child Sexual Abuse or Exploitation Materials Act, and its Implementing Rules and Regulations. It examined four dimensions of local capacity: legal, administrative, relational, and fiscal. It also assessed the costs and investments required to implement and sustain local anti-OSAEC programs.

The study employed a mixed-method approach in Manila, Dasmariñas City in Cavite, and Llorente in Eastern Samar. The qualitative component comprised 15 key informant interviews and 10 focus group discussions with representatives from …

Happi Phase Ii Learning And Other Learning-Related Activities In Select Sites In The Philippines: Final Report, Zaldy Collado, Homer Yabut, Jerome Cleofas, Rowalt Alibudbud, Marites Tiongco, Pilar Ramos-Jimenez, Cecile Lantican, Marion Navarro, Maria Margareth Geluz, Gi-Ann Concepcion

SDRC Projects

This final report examines barriers to and factors that support human papillomavirus (HPV) vaccination in selected Philippine sites. It presents findings from key informant interviews and an audit of HPV vaccination communication materials and practices. The report proposes a policy framework to improve program implementation and expand vaccination coverage, including outreach to out-of-school girls.

The study examined how health systems, community engagement, and coordination across sectors shape HPV vaccination in the Philippines. Guided by a Theory of Change, the research team interviewed 137 key informants representing 163 stakeholder positions across national and local levels. The study covered 10 sites with …

Testing Alternative Versions Of The Fama-French Five-Factor Model: Evidence From Chinese Non-Financial Firms, Su Zhang

Plymouth Business School Theses

This research investigates the applicability and effectiveness of the Fama-French model within the Chinese stock market. While the Fama-French model is extensively used in asset pricing, its performance in emerging markets like China remains insufficiently explored. The Chinese stock market is characterized by unique features such as the prevalence of state-owned enterprises, a high number of retail investors, market segmentation, government intervention. These elements may impact the traditional Fama-French model's explanatory power, necessitating modifications and adjustments.The thesis tests the Fama-French five-factor model's applicability in China's A-share market (2006–2023) by adding a risk (RISK) factor. Using monthly data, it constructs portfolios …

New York Environmental Legislation In 2024, Michael B. Gerrard

Faculty Scholarship

In 2024, New York State enacted several laws aimed either at dealing with the adverse impacts of climate change such as extreme heat and increasingl severe and frequent storms or at helping to mitigate greenhouse gas emissions to avoid or minimize the adverse impacts. The final bill signed by Governor Kathy Hochul in 2024 was the most high-profile of these laws — the Climate Change Superfund Act, which will require fossil fuel companies to pay for adaptation measures.

Other new laws last year related to extreme heat in schools and prisons, food waste, expanded polystyrene containers, community gardens, and native …

The Role Of Sociocultural Factors In Saudi Consumers’ Attitudes Towards Social Media Advertising And Purchase Intentions Within The Clothing Industry, Lamis Aldawish

Plymouth Business School Theses

Social media advertising (SMA) significantly influences consumer purchase behaviour in Saudi Arabia, particularly in the fashion industry. However, Saudi consumers’ attitudes towards fashion advertisements on social media are shaped by sociocultural factors, given the country’s deep-rooted cultural and Islamic principles. With Saudi Vision 2030 driving societal changes and transforming the fashion industry, younger generations (under 45 years) are developing different attitudes compared to older generations (over 45 years). Although notable changes have occurred, existing literature lacks sufficient exploration of how sociocultural factors influence Saudi consumers’ attitudes towards social media advertising and their intentions to purchase clothing. Gaining insight into these …

For Whose Benefit Is The Freedom Of Speech?, Vincent A. Blasi

Faculty Scholarship

In United States v. Alvarez, the US Supreme Court ruled that an official of a water district who introduced himself to his constituents by falsely stating in a public meeting that he had earned the Congressional Medal of Honor had a First Amendment right to make that demonstrably untrue claim. Audience members misled by the statement might well be considered to have a First Amendment interest in not being directly and knowingly lied to in that way. Other members of the community might be thought to have a First Amendment interest in public officials such as Xavier Alvarez telling …

Graphene Biosensors For The Detection Of Alzheimer's And Cardiovascular Disease Protein Biomarkers, Theodore Bungon

School of Engineering, Computing and Mathematics Theses

Dementia describes a syndrome or group of symptoms associated with the decline of brain functioning such as loss of memory, mood, judgment, language, thinking speed, mental sharpness and other mental abilities. Different classes of Dementia are Jakob Disease, Korsakoff Syndrome, Vascular dementia, Parkinson’s disease, Huntington disease, young onset dementia, Alzheimer’s disease (AD) etc. At present, AD is irreversible and incurable, and it worsens overtime, it is neurodegenerative, caused by the accumulation or build-up of abnormal protein in the brain. It is a type of dementia that is common, affecting about 60% of patients with dementia diagnosis. Globally, over 54 million …

Generated Image Quality Assurance And Generative Adversarial Network Augmentation For Machine Learning, Dirk Holscher

School of Engineering, Computing and Mathematics Theses

The advancement of machine learning is heavily dependent on the quality of data used for training models. This thesis explores the enhancement of data quality and augmentation techniques using Generative Adversarial Networks (GANs), pecifically focusing on the Pix2Pix architecture. The research addresses the critical challenges of improving image quality, optimizing hyperparameters, and detecting fake images generated by GANs, aiming to enhance the robustness and eliability of machine learning models. This work presents a collection of metrics for measuring data quality for both structured and unstructured data. Furthermore, it defines data quality for machine learning, divided into three quality levels. Advanced …

Bayesian Multilevel Models For Survival Data From Multiple Centres, Lingling Zhang

School of Engineering, Computing and Mathematics Theses

Recently, there has been an increasing use of flexible parametric survival models to analyse survival outcomes. Much of our work is based on a survival analysis model in which the baseline hazard function is modelled flexibly using an M-spline. Motivated by the analysis of a large kidney transplant data set supplied by NHSBT (NHS Blood and Transplant), we extend this baseline hazard function flexibility in various ways to allow a random effect for each transplant centre.After providing an introduction to survival analysis, we present an exploratory analysis of the NHSBT kidney transplant data. Next, we analyse the data using Bayesian …

A Protocol Agnostic Electrically Transparent Distributed Switch Fabric For It/Ot Testbenches, Wesley Andrews

School of Engineering, Computing and Mathematics Theses

This thesis presents the design and development of custom hardware, based upon RF (Radio Frequency) MEMS (Micro Electro-Mechanical System) switch technology and a grounded, conductor backed, edge coupled, coplanar waveguide transmission line design to facilitate protocol-agnostic, electrically transparent switching of different hard-wire protocols, such as Ethernet, USB (Universal Serial Bus), CAN (Controller Area Network) bus, RF and more within an IT/OT (Informational/Operational Technology) environment. This is achieved through a distributed, decentralised switch fabric composed of individual switch fabric nodes, which achieves fully reconfigurable IT/OT environments at the electrical level. Traditional IT/OT environments consist of diverse systems involving computers, network switches, …