Physical Sciences and Mathematics Commons^™

Adversarial Learning For Coordinate Regression Through K-Layer Penetrating Representation, Mengxi Jiang, Yulei Sui, Yunqi Lei, Xiaofei Xie, Cuihua Li, Yang Liu, Ivor W. Tsang

Research Collection School Of Computing and Information Systems

Adversarial attack is a crucial step when evaluating the reliability and robustness of deep neural networks (DNNs) models. Most existing attack approaches apply an end-to-end gradient update strategy to generate adversarial examples for a classification or regression problem. However, few of them consider the non-differentiable DNN models (e.g., coordinate regression model) that prevent end-to-end backpropagation resulting in the failure of gradient calculation. In this paper, we present a new adversarial example generation approach for both untargeted and targeted attacks on coordinate regression models with non-differentiable operations. The novelty of our approach lies in a k-layer penetrating representation, on which we …

Efficient Multiplicative-To-Additive Function From Joye-Libert Cryptosystem And Its Application To Threshold Ecdsa, Haiyang Xue, Ho Man Au, Mengling Liu, Yin Kwan Chan, Handong Cui, Xiang Xie, Hon Tsz Yuen, Chengru Zhang

Research Collection School Of Computing and Information Systems

Threshold ECDSA receives interest lately due to its widespread adoption in blockchain applications. A common building block of all leading constructions involves a secure conversion of multiplicative shares into additive ones, which is called the multiplicative-to-additive (MtA) function. MtA dominates the overall complexity of all existing threshold ECDSA constructions. Specifically, O(n2) invocations of MtA are required in the case of n active signers. Hence, improvement of MtA leads directly to significant improvements for all state-of-the-art threshold ECDSA schemes.In this paper, we design a novel MtA by revisiting the Joye-Libert (JL) cryptosystem. Specifically, we revisit JL encryption and propose a JL-based …

Developing Policies For Digital Twin Data Quality And Security Controls, Ahmad Abdelbaset Hassan

Theses

This thesis is concerned with the data quality and security of the digital twin and how it is going to impact its adoption, trustworthiness, and potential for real-world applications. By addressing the potential vulnerabilities and ensuring the integrity of data, this research aims to contribute to the development of robust and trustworthy digital twin policies that to complement the existing international standards across different domains. Moreover, it underscores the important need to establish robust policies to ensure the successful and secure deployment of digital twins across industries. Previous research, while valuable, may not have fully addressed the critical interplay between …

Advancing Crispr-Based Solutions For Covid-19 Diagnosis And Therapeutics, Roaa Hadi, Abhishek Poddar, Shivakuma Sonnaila, Venkata Suryanarayana Murthy Bhavaraju, Shilpi Agrawal

Chemistry & Biochemistry Faculty Publications and Presentations

Since the onset of the COVID-19 pandemic, a variety of diagnostic approaches, including RT-qPCR, RAPID, and LFA, have been adopted, with RT-qPCR emerging as the gold standard. However, a significant challenge in COVID-19 diagnostics is the wide range of symptoms presented by patients, necessitating early and accurate diagnosis for effective management. Although RT-qPCR is a precise molecular technique, it is not immune to false-negative results. In contrast, CRISPR-based detection methods for SARS-CoV-2 offer several advantages: they are cost-effective, time-efficient, highly sensitive, and specific, and they do not require sophisticated instruments. These methods also show promise for scalability, enabling diagnostic tests. …

Optimizing Uncertainty Estimation On Scientific Visualizations Using Learning Models, Clara Almeida, David Guerrero-Pantoja, Erik Pautsch, Silvio Rizzi, George Thiruvathukal, Maria Pantoja

Computer Science: Faculty Publications and Other Works

Scientific visualizations (SciVis) translate numerical and spatial data as images, enabling scientists to better understand the phenomena described by the data and gain insights that may be overlooked by statistical methods alone. In recent years, machine learning, particularly Deep Learning (DL), has driven significant advancements in SciVis. By combining classical numerical techniques with DL methods, we can create models that potentially achieve the accuracies needed for real-world industrial applications and simulations. However, DL models often exhibit undue confidence, particularly when faced with inputs that differ significantly from the data they were trained on, known as out-of-distribution (OOD) inputs. This can …

Puccinellia (Puccinellia Ciliata) For Dryland Salinity Management, Department Of Primary Industries And Regional Development, Western Australia

Natural resources factsheets

Puccinellia (Puccinellia ciliata) is a potentially valuable fodder on waterlogged saline land. Puccinellia, Distichlis and marine couch are the most salt tolerant of the commercially available grasses. It is best suited to areas with more than 400mm annual rainfall and where the watertable is shallow over summer.

The Department of Primary Industries and Regional Development recommends that any dryland salinity management is part of a whole farm, and preferably a whole catchment, water management plan.

Roaded Catchments In Wa, Department Of Primary Industries And Regional Development, Western Australia

Natural resources factsheets

Roaded catchments are earthwork structures designed to increase the amount of run-off from the catchment above a receiving farm dam.

Roaded catchments are lined with clay and compacted to make a smooth surface that reduces infiltration and increases run-off. Roaded catchments should discharge into a silt trap and then into a dam (excavated earth tank). This combined structure should be fenced to exclude livestock.

The Department of Primary Industries and Regional Development recommends that roaded catchments are part of an integrated water and salinity management program.

The information in this page is only a guide – seek expert advice before …

Managing Subsurface Water In Wa, Department Of Primary Industries And Regional Development, Western Australia

Natural resources factsheets

The Department of Primary Industries and Regional Development (DPIRD) recommends that subsurface water management is part of an integrated water and salinity program. An integrated program can help to lower water tables and alleviate problems with waterlogging, rising salinity, and infrastructure damage.

Excavated Tanks (Farm Dams), Department Of Primary Industries And Regional Development, Western Australia

Natural resources factsheets

Guidelines and information for planning, legal requirements, design, construction and management of excavated tanks for water storage on farm.

Excavated tanks (farm dams) provide effective water storage wherever surface water run-off can be harvested for livestock, crop spraying, irrigation and domestic use on rural properties.

Excavated tanks for rainfall storage are common in the southern agricultural areas of WA.

Well-designed and constructed dams are farm assets, safe and require little maintenance.

Poor design and construction can lead to poor water harvesting and storage, excessive costs, downstream erosion and risks to property and people.

DPIRD recommends that excavated tanks should be …

Pumping Groundwater For Salinity Control In Wa, Department Of Primary Industries And Regional Development, Western Australia

Natural resources factsheets

DPIRD recommends that groundwater pumping is used as part of an integrated water and salinity management program.

Groundwater pumping can lower local water tables, recover land for production and reduce the spread of local salinity or to protect high value assets at risk, such as townsites, roads or railway lines. Pump water can be discharged to safe disposal points or used on farm for desalination.

Improving Lambing Rates And Changing Turnoff Times To Reduce Greenhouse Gas Emissions, Mandy Curnow, Sud Kharel, Claire Macleay

Animal production and livestock research reports

Running a flock of sheep and raising lambs generates both direct (enteric fermentation) and indirect (lime, fertiliser, fuel, production of feed) greenhouse gas (GHG) emissions. There are ways to reduce emissions through sheep reproduction.

Greenhouse Gas Accounting For Western Australian Agriculture 2022, Johnny Machon, Mandy Curnow, Christophe D'Abbadie

Climate Science Published Reports

This report presents new modelling of the greenhouse gas emissions attributable to Western Australia’s agricultural industries and provides an overview of sequestration at the state level. The report improves and expands on the estimates from previous technical reports by the Department of Primary Industries and Regional Development (DPIRD), most notably Curnow et al. (2022). The report presents a new, much broader picture, capturing not only emissions from agricultural activities in Western Australia (WA), but also the emissions associated with the production of agricultural inputs and the sequestration that occurs on agriculturally managed lands. This broader picture of emissions aligns with …

Lithicone-Protected Lithium Metal Anodes For Lithium Metal Batteries With Nickel-Rich Cathode Materials, Ridwan A. Ahmed, Kevin V. Carballo, Krishna P. Koirala, Qian Zhao, Peiyuan Gao, Ju-Myung Kim, Cassidy S. Anderson, Xiangbo Meng, Chongmin Wang, Ji-Guang Zhang, Wu Xu

Mechanical Engineering Faculty Publications and Presentations

The high energy density advantage of lithium (Li) metal batteries (LMBs) makes them increasingly desirable; however, problems such as strong reactivity and dendrite growth of Li metal anode limit their practical uses. In this work, a novel Li-containing glycerol (LiGL) or lithicone protection layer on a 50 μm thick Li metal anode is employed for improving the performance of LMBs. This LiGL layer was accurately deposited via a molecular layer deposition (MLD) process at 150°C, using lithium tert-butoxide and glycerol as precursors. The as-formed LiGL coating layer is highly tunable in its thickness by simply adjusting MLD cycles and shows …

Unital Kadison-Schwarz Type Of Maps, Hajir Khalid Al Zadjali

Theses

Quantum entanglement is an important phenomenon in quantum information theory; its uses range from quantum teleportation and quantum cryptography to solving classical communication problems. One of the issues concerning entanglement is the classification of mixed states into separable states. Positive maps define affine mappings between sets of states of C∗algebras, and have been used to develop a separability criterion for quantum states. Completely positive maps are insensitive to entanglement detection, a weaker notion of positivity is thus needed to detect them. This condition that is stronger than positivity but weaker than complete positivity is given by the Kadison-Schwarz (KS) inequality. …

Quantum Markov Chains Related To Certain Lattice Models, Ali Ghaleb Alaali

Theses

A central open problem in quantum probability is the establishment of a general theory of Markov fields. This thesis contributes to the general theory by introducing quantum probability and applies it via the construction of quantum Markov chains on different hierarchical lattices (Cayley trees). This thesis examines two types of trees one of which corresponds to the Ising-XY-Model on a Cayley tree of order two which then the existence of Markov chains can be utilized to detect phase transitions. The other discussed model is a (1,3) tree in which the XY-model is considered to determine the existence of a quantum …

Ai-Powered Pedagogy: Revolutionizing Students’ Learning Experiences Through Integration Of Ai Technologies, Amna Awad Alsaedi

Theses

This research examines the integration of Artificial Intelligence (AI) within the educational sector with the aim of enhancing student learning outcomes. AI offers tailored learning experiences, interactive educational content, prompt feedback, and access to diverse learning resources. Nonetheless, challenges including addiction, costliness, privacy infringement, bias, ethical dilemmas, market competition, and moral considerations require resolution. Research particularly delves into the utilization of chatbots, and algorithms designed to simulate human interactions and generate text resembling human speech. Educational applications powered by AI have the potential to heighten student engagement, comprehension, and academic performance by substituting traditional assignments with concise, information-rich lessons. Furthermore, …

Corrosion Behavior Of 304 Stainless Steel In H2so4/Nacl In The Absence And Presence Of Molybdate And Tungstate, Asma Yousef Alkaabi

Theses

In this study, the corrosion behavior of 304 austenitic stainless steel was investigated in H₂SO₄ solutions, H₂SO₄ as a function of Cl^- ions concentration, H₂SO₄ as a function of MoO4^2- ions concentration, and H₂SO₄ + Cl^- ions as a function of MoO₄^2- ions concentration. Moreover, the synergistic inhibition of MoO4^2- and WO₄^2- ions was investigated in 0.1 M H₂SO₄, with the solution pH adjusted to 3.0. The findings show that the corrosion current density (i …

Assessing The Spatiotemporal Variability Of Aerosol Optical Depth And Particle Sizes In The Uae Atmosphere Using Remotely Sensed Data, Bashayer Ali Alzahmi

Theses

This thesis investigates atmospheric aerosol concentrations, with a particular focus on the United Arab Emirates (UAE). The study maps the spatial and temporal variations in Aerosol Optical Depth (AOD) and characterizes aerosols using high-resolution Moderate Resolution Imaging Spectroradiometer (MODIS) products. The primary objective of this research is to enhance the understanding of aerosol dynamics in the UAE by distinguishing between natural and anthropogenic aerosols through the Angström Exponent (AE) and assessing their impacts on local climate, air quality, and public health. The study employs Remote Sensing (RS) techniques to conduct detailed environmental assessments, analyzing AOD variability across daily, monthly, seasonal, …

Exclusion Tests For Approximating All Solutions Of Nonlinear Equations, Yasmeen Omar Hamida

Theses

The exclusion test is a mathematical technique used to isolate and approximate the solutions of nonlinear equations within a specified interval. This thesis aims to construct a new test leveraging the principles of the exclusion test to achieve faster convergence to the zeros of any given interval within a compact domain. By selecting an alternative objective function and employing nonlinear equations, the proposed method significantly reduces the number of subintervals required for accurate approximation. The key finding of this research is the development of a more efficient algorithm that enhances the speed of convergence. The implications of these findings are …

Effect Of Virtual Reality Technology On Computer Science/Engineering Based Laboratories Education – A Case Study, Mariam A. Al Nuaimi

Theses

Virtual reality (VR) is becoming increasingly popular and essential in education as institutions strive to incorporate technology in education. In Engineering and Computer Science, students face difficulties in comprehending many abstract complex concepts that are typically studied in Labs. This situation worsens with hardware failure and the lack of pedagogical tools. The main goal of this paper is to investigate the impact of VR environments on learning advanced and complex STEM concepts. We also explore the impact of different human-computer interaction (HCI) techniques including gamification on the learning experience of STEM students. To measure such impact, we developed a VR …

Effect Of Virtual Reality Technology On Computer Science/Engineering Based Laboratories Education – A Case Study, Saeed Salem Al Shebli

Theses

The rapid growth in mobile applications raises critical concerns about the security of apps and users' privacy, especially in permission control. Mobile apps access sensitive information of users, and the current cybersecurity landscape faces a huge challenge in ensuring the least required permissions are granted. This research focuses on designing an advanced permission recommendation system that couples the strengths of Natural Language Processing (NLP) and Machine Learning (ML) in solving most of the existing gaps in permission management, thus guiding which permissions are mostly needed by Android applications.

The research thus follows a multi-classification approach, integrating state-of-the-art ML techniques with …

On Fractional Dunkl-Type Laplacian, Saba Ibrahim Aldan

Theses

This thesis presents a comprehensive study of the fractional Dunkl-type Laplacian operator (−∥x∥ Δ��)^σ for 0<σ<1, through four equivalent key characterizations: the heat semigroup characterization, the pointwise characterization, the spherical mean characterization, and through an extension theorem. These characterizations offer different perspectives on the operator ∥x∥Δ_��, generalizing known results for the Euclidean Laplace operator by incorporating reflection symmetries, making it highly relevant in the harmonic analysis of root systems.

By exploring these characterizations, the thesis highlights the deep connections between the fractional operator (−∥x∥Δ_��)^σ and harmonic analysis, as well as its broader applicability in mathematical physics and partial differential equations.

A Secure And Effective Framework For Key Concept Mining From Educational Content Using Large Language Models, Ashika Sameem Abdul Rasheed

Theses

This thesis examines the use of Large Language Models (LLMs) in education, with a focus on improving performance and implementing strong security measures. The research has two main goals, namely, the development of an effective lecture summarization technique using LLMs and identifying and addressing security vulnerabilities in LLM applications according to OWASP (Open Web Application Security Project) guidelines. For the former goal, we have proposed an effective framework for fine-tuning LLMs using real lecture datasets and compared the performance of different LLMs. For the latter goal, we conducted a thorough review of the application dataflow of the proposed framework and …

Improving Conversational Recommender System Via Contextual And Time-Aware Modeling With Less Domain-Specific Knowledge, Lingzhi Wang, Shafiq Joty, Wei Gao, Xingshan Zeng, Kam-Fai Wong

Research Collection School Of Computing and Information Systems

Conversational Recommender Systems (CRS) has become an emerging research topic seeking to perform recommendations through interactive conversations, which generally consist of generation and recommendation modules. Prior work on CRS tends to incorporate more external and domain-specific knowledge like item reviews to enhance performance. Despite the fact that the collection and annotation of the external domain-specific information needs much human effort and degenerates the generalizability, too much extra knowledge introduces more difficulty to balance among them. Therefore, we propose to fully discover and extract the internal knowledge from the context. We capture both entity-level and contextual-level representations to jointly model user …

A Comprehensive Survey On Relation Extraction: Recent Advances And New Frontiers, Xiaoyan Zhao, Yang Deng, Min Yang, Lingzhi Wang, Rui Zhang, Hong Cheng, Wai Lam, Ying Shen, Ruifeng Xu

Research Collection School Of Computing and Information Systems

Relation extraction (RE) involves identifying the relations between entities from underlying content. RE serves as the foundation for many natural language processing (NLP) and information retrieval applications, such as knowledge graph completion and question answering. In recent years, deep neural networks have dominated the field of RE and made noticeable progress. Subsequently, the large pre-trained language models (PLMs) have taken the state-of-the-art RE to a new level. This survey provides a comprehensive review of existing deep learning techniques for RE. First, we introduce RE resources, including datasets and evaluation metrics. Second, we propose a new taxonomy to categorize existing works …

Cas: Fusing Dnn Optimization & Adaptive Sensing For Energy-Efficient Multi-Modal Inference, Dulanga Weerakoon, Vigneshwaran Subbaraju, Joo Hwee Lim, Archan Misra

Research Collection School Of Computing and Information Systems

Intelligent virtual agents are used to accomplish complex multi-modal tasks such as human instruction comprehension in mixed-reality environments by increasingly adopting richer, energy-intensive sensors and processing pipelines. In such applications, the context for activating sensors and processing blocks required to accomplish a given task instance is usually manifested via multiple sensing modes. Based on this observation, we introduce a novel Commit-and-Switch ( CAS ) paradigm that simultaneously seeks to reduce both sensing and processing energy. In CAS , we first commit to a low-energy computational pipeline with a subset of available sensors. Then, the task context estimated by this pipeline …

Strength Lies In Differences! Improving Strategy Planning For Non-Collaborative Dialogues Via Diversified User Simulation, Tong Zhang, Chen Huang, Yang Deng, Hongru Liang, Jia Liu, Zujie Wen, Wenqiang Lei, Tat-Seng Chua

Research Collection School Of Computing and Information Systems

We investigate non-collaborative dialogue agents, which are expected to engage in strategic conversations with diverse users, for securing a mutual agreement that leans favorably towards the system’s objectives. This poses two main challenges for existing dialogue agents: 1) The inability to integrate user-specific characteristics into the strategic planning, and 2) The difficulty of training strategic planners that can be generalized to diverse users. To address these challenges, we propose TRIP to enhance the capability in tailored strategic planning, incorporating a user-aware strategic planning module and a population-based training paradigm. Through experiments on benchmark non-collaborative dialogue tasks, we demonstrate the effectiveness …

Unlocking Markets: A Multilingual Benchmark To Cross-Market Question Answering, Yifei Yuan, Yang Deng, Anders Sogaard, Mohammad Alliannejadi

Research Collection School Of Computing and Information Systems

Users post numerous product-related questions on e-commerce platforms, affecting their purchase decisions. Product-related question answering (PQA) entails utilizing product-related resources to provide precise responses to users. Wepropose a novel task of Multilingual Crossmarket Product-based Question Answering (MCPQA) and define the task as providing answers to product-related questions in a main marketplace by utilizing information from another resource-rich auxiliary marketplace in a multilingual context. We introduce a largescale dataset comprising over 7 million questions from 17 marketplaces across 11 languages. We then perform automatic translation on the Electronics category of our dataset, naming it as McMarket. We focus on two subtasks: …

Ask-Before-Plan : Proactive Language Agents For Real-World Planning, Xuan Zhang, Yang Deng, Zifeng Ren, See-Kiong Ng, Tat-Seng Chua

Research Collection School Of Computing and Information Systems

The evolution of large language models (LLMs) has enhanced the planning capabilities of language agents in diverse real-world scenarios. Despite these advancements, the potential of LLM-powered agents to comprehend ambiguous user instructions for reasoning and decision-making is still under exploration. In this work, we introduce a new task, Proactive Agent Planning, which requires language agents to predict clarification needs based on user-agent conversation and agent-environment interaction, invoke external tools to collect valid information, and generate a plan to fulfill the user's demands. To study this practical problem, we establish a new benchmark dataset, Ask-before-Plan. To tackle the deficiency of LLMs …

Thoughts To Target: Enhance Planning For Target-Driven Conversation, Zhonghua Zheng, Lizi Liao, Yang Deng, Ee-Peng Lim, Minlie Huang, Liqiang Nie

Research Collection School Of Computing and Information Systems

In conversational AI, large-scale models excel in various tasks but struggle with target-driven conversation planning. Current methods, such as chain-of-thought reasoning and tree-search policy learning techniques, either neglect plan rationality or require extensive human simulation procedures. Addressing this, we propose a novel two-stage framework, named EnPL, to improve the LLMs’ capability in planning conversations towards designated targets, including (1) distilling natural language plans from target-driven conversation corpus and (2) generating new plans with demonstration-guided in-context learning. Specifically, we first propose a filter approach to distill a high-quality plan dataset, ConvPlan1. With the aid of corresponding conversational data and support from …