Physical Sciences and Mathematics Commons^™

Electron-Transfer Mediated Photo-Switching In Nanoparticles, Liaoran Cao

All Dissertations

Previous publications from our lab demonstrated viable approaches to design a photo-switching nanoparticle with arguably superior brightness and photostability resulting in an improved resolution in localization-based microscopy, as compared to other photo-switching dyes or particles. As a follow-up, this dissertation is focused on two major tasks: first, developing nanoparticles with better photo-switching properties for super-resolution imaging; second, trying to achieve a better physical picture of the mechanisms involved in photo-switching, including polaron dynamics, charge transfer, and energy transfer. A new class of photo-switchable nanoparticles was developed by blending conjugated polymer semiconductor with fullerene-based electron acceptors, and further blending with polystyrene …

Atom Economical Reactions Of Terpenoids And Post-Consumer Plastics With Sulfur, Charini Maladeniya

All Dissertations

The production of Ordinary Portland Cement (OPC) and its uses have a complicated environmental impact, which is influenced by infrastructure development and building operations as well as CO₂ emissions, which account for 7% of all worldwide CO₂ emissions. Due to the increasing population, production is still ongoing. The search for cement and construction materials produced with zero to low CO₂ emissions is therefore continuous. Finding recyclable, CO₂ gas-free biocomposites with high sulfur content that can rival the mechanical properties of popular building supplies like Portland cement is the primary objective of the research discussed in this …

Motion Synthesis And Control For Autonomous Agents Using Generative Models And Reinforcement Learning, Pei Xu

All Dissertations

Imitating and predicting human motions have wide applications in both graphics and robotics, from developing realistic models of human movement and behavior in immersive virtual worlds and games to improving autonomous navigation for service agents deployed in the real world. Traditional approaches for motion imitation and prediction typically rely on pre-defined rules to model agent behaviors or use reinforcement learning with manually designed reward functions. Despite impressive results, such approaches cannot effectively capture the diversity of motor behaviors and the decision making capabilities of human beings. Furthermore, manually designing a model or reward function to explicitly describe human motion characteristics …

Recovering Coefficients Of Second-Order Hyperbolic And Plate Equations Via Finite Measurements On The Boundary, Scott Randall Scruggs

All Dissertations

Abstract In this dissertation, we consider the inverse problem for a second-order hyperbolic equation of recovering n + 3 unknown coefficients defined on an open bounded domain with a smooth enough boundary. We also consider the inverse problem of recovering an unknown coefficient on the Euler- Bernoulli plate equation on a lower-order term again defined on an open bounded domain with a smooth enough boundary. For the second-order hyperbolic equation, we show that we can uniquely and (Lipschitz) stably recover all these coefficients from only using half of the corresponding boundary measurements of their solutions, and for the plate equation, …

Acceleration Methods For Nonlinear Solvers And Application To Fluid Flow Simulations, Duygu Vargun

All Dissertations

This thesis studies nonlinear iterative solvers for the simulation of Newtonian and non- Newtonian fluid models with two different approaches: Anderson acceleration (AA), an extrapolation technique that accelerates the convergence rate and improves the robustness of fixed-point iterations schemes, and continuous data assimilation (CDA) which drives the approximate solution towards coarse data measurements or observables by adding a penalty term.

We analyze the properties of nonlinear solvers to apply the AA technique. We consider the Picard iteration for the Bingham equation which models the motion of viscoplastic materials, and the classical iterated penalty Picard and Arrow-Hurwicz iterations for the incompressible …

Generative Pre-Trained Transformers (Gpt) And Space Health: A Potential Frontier In Astronaut Health During Exploration Missions, Ethan Waisberg, Joshua Ong, Mouayad Masalkhi, Nasif Zaman, Sharif Amit Kamran, Andrew G Lee

Faculty, Staff and Student Publications

In anticipation of space exploration where astronauts are traveling away from Earth, and for longer durations with an increasing communication lag, artificial intelligence (AI) frameworks such as large language learning models (LLMs) that can be trained on Earth can provide real-time answers. This emerging technology may be helpful for acute medical emergencies, particularly in austere and distant space environments. In this manuscript, we provide an overview of generative pre-trained transformer (GPT) technology, a rapidly emerging AI technology, and implications, considerations, and limitations of such technology for space health.

Atomically Dispersed Cu-N3 On Hollow Spherical Carbon Nitride For Acetaminophen Degradation: Generation Of 1o2 From H2o2, Shuoyuan Tian, Yu Yin, Mengxuan Liu, Lei Shi, Shu Zhang, Abdul Hannan Asif, Xuan Li, Mengqiong Liu, Xiaoguang Duan, Shaobin Wang, Hongqi Sun

Research outputs 2022 to 2026

Discharge of recalcitrant pharmaceuticals into aquatic systems has caused severe impacts on public health and ecosystem. Advanced oxidation processes (AOPs) are effective for eliminating these refractory pollutants, for which single-atom catalysts (SACs) become the state-of-the-art materials owing to the maximized exposure of active metal sites. In this work, hollow spherical graphitic carbon nitride (hsCN) was fabricated to incorporate copper species to develop Fenton-like catalysts for acetaminophen (ACT) removal. Through pyrolysis of supramolecular assemblies derived from melamine-Cu complex and cyanuric acid, single atom Cu-N3 sites were anchored on hsCN by N-coordination to obtain SACu-hsCN. In virtue …

Ultrasonic Study Of Water Adsorbed In Nanoporous Glasses, Jason Ogbebor, John J. Valenza, Peter I. Ravikovitch, Ashoka Karunarathne, Giovanni Muraro, Maxim Lebedev, Boris Gurevich, Alexei F. Khalizov, Gennady Y. Gor

Research outputs 2022 to 2026

Thermodynamic properties of fluids confined in nanopores differ from those observed in the bulk. To investigate the effect of nanoconfinement on water compressibility, we perform water sorption experiments on two nanoporous glass samples while concomitantly measuring the speed of longitudinal and shear ultrasonic waves in these samples. These measurements yield the longitudinal and shear moduli of the water-laden nanoporous glass as a function of relative humidity that we utilize in the Gassmann theory to infer the bulk modulus of the confined water. This analysis shows that the bulk modulus (inverse of compressibility) of confined water is noticeably higher than that …

Survey On Sentiment Analysis: Evolution Of Research Methods And Topics, Jingfeng Cui, Zhaoxia Wang, Seng-Beng Ho, Erik Cambria

Research Collection School Of Computing and Information Systems

Sentiment analysis, one of the research hotspots in the natural language processing field, has attracted the attention of researchers, and research papers on the field are increasingly published. Many literature reviews on sentiment analysis involving techniques, methods, and applications have been produced using different survey methodologies and tools, but there has not been a survey dedicated to the evolution of research methods and topics of sentiment analysis. There have also been few survey works leveraging keyword co-occurrence on sentiment analysis. Therefore, this study presents a survey of sentiment analysis focusing on the evolution of research methods and topics. It incorporates …

Decoding The Underlying Meaning Of Multimodal Hateful Memes, Ming Shan Hee, Wen Haw Chong, Roy Ka-Wei Lee

Research Collection School Of Computing and Information Systems

Recent studies have proposed models that yielded promising performance for the hateful meme classification task. Nevertheless, these proposed models do not generate interpretable explanations that uncover the underlying meaning and support the classification output. A major reason for the lack of explainable hateful meme methods is the absence of a hateful meme dataset that contains ground truth explanations for benchmarking or training. Intuitively, having such explanations can educate and assist content moderators in interpreting and removing flagged hateful memes. This paper address this research gap by introducing Hateful meme with Reasons Dataset (HatReD), which is a new multimodal hateful meme …

Transferable Curricula Through Difficulty Conditioned Generators, Sidney Tio, Pradeep Varakantham

Research Collection School Of Computing and Information Systems

Advancements in reinforcement learning (RL) have demonstrated superhuman performance in complex tasks such as Starcraft, Go, Chess etc. However, knowledge transfer from Artificial "Experts" to humans remain a significant challenge. A promising avenue for such transfer would be the use of curricula. Recent methods in curricula generation focuses on training RL agents efficiently, yet such methods rely on surrogate measures to track student progress, and are not suited for training robots in the real world (or more ambitiously humans). In this paper, we introduce a method named Parameterized Environment Response Model (PERM) that shows promising results in training RL agents …

Learning To Send Reinforcements: Coordinating Multi-Agent Dynamic Police Patrol Dispatching And Rescheduling Via Reinforcement Learning, Waldy Joe, Hoong Chuin Lau

Research Collection School Of Computing and Information Systems

We address the problem of coordinating multiple agents in a dynamic police patrol scheduling via a Reinforcement Learning (RL) approach. Our approach utilizes Multi-Agent Value Function Approximation (MAVFA) with a rescheduling heuristic to learn dispatching and rescheduling policies jointly. Often, police operations are divided into multiple sectors for more effective and efficient operations. In a dynamic setting, incidents occur throughout the day across different sectors, disrupting initially-planned patrol schedules. To maximize policing effectiveness, police agents from different sectors cooperate by sending reinforcements to support one another in their incident response and even routine patrol. This poses an interesting research challenge …

Camper: An Effective Framework For Privacy-Aware Deep Entity Resolution, Yuxiang Guo, Lu Chen, Zhengjie Zhou, Baihua Zheng, Ziquan Fang, Zhikun Zhang, Yuren Mao, Yunjun Gao

Research Collection School Of Computing and Information Systems

Entity Resolution (ER) is a fundamental problem in data preparation. Standard deep ER methods have achieved state-of-the-art efectiveness, assuming that relations from diferent organizations are centrally stored. However, due to privacy concerns, it can be difcult to centralize data in practice, rendering standard deep ER solutions inapplicable. Despite eforts to develop rule-based privacy-preserving ER methods, they often neglect subtle matching mechanisms and have poor efectiveness as a result. To bridge efectiveness and privacy, in this paper, we propose CampER, an efective framework for privacy-aware deep entity resolution. Specifcally, we frst design a training pair self-generation strategy to overcome the absence …

Multi-View Graph Contrastive Learning For Solving Vehicle Routing Problems, Yuan Jiang, Zhiguang Cao, Yaoxin Wu, Jie Zhang

Research Collection School Of Computing and Information Systems

Recently, neural heuristics based on deep learning have reported encouraging results for solving vehicle routing problems (VRPs), especially on independent and identically distributed (i.i.d.) instances, e.g. uniform. However, in the presence of a distribution shift for the testing instances, their performance becomes considerably inferior. In this paper, we propose a multi-view graph contrastive learning (MVGCL) approach to enhance the generalization across different distributions, which exploits a graph pattern learner in a self-supervised fashion to facilitate a neural heuristic equipped with an active search scheme. Specifically, our MVGCL first leverages graph contrastive learning to extract transferable patterns from VRP graphs to …

Decompiling X86 Deep Neural Network Executables, Zhibo Liu, Yuanyuan Yuan, Shuai Wang, Xiaofei Xie, Lei Ma

Research Collection School Of Computing and Information Systems

Due to their widespread use on heterogeneous hardware devices, deep learning (DL) models are compiled into executables by DL compilers to fully leverage low-level hardware primitives. This approach allows DL computations to be undertaken at low cost across a variety of computing platforms, including CPUs, GPUs, and various hardware accelerators. We present BTD (Bin to DNN), a decompiler for deep neural network (DNN) executables. BTD takes DNN executables and outputs full model specifications, including types of DNN operators, network topology, dimensions, and parameters that are (nearly) identical to those of the input models. BTD delivers a practical framework to process …

Fintech Data Infrastructure For Esg Disclosure Compliance, Randall E. Duran, Peter Tierney

Research Collection School Of Computing and Information Systems

Regulations related to the disclosure of environmental, governance, and social (ESG) factors are evolving rapidly and are a major concern for financial compliance worldwide. Information technology has the potential to reduce the effort and cost of ESG disclosure compliance. However, comprehensive and accurate ESG data are necessary for disclosures. Currently, the availability and quality of underlying data for ESG disclosures vary widely and are often deficient. The process involved with obtaining ESG data is also often inefficient and prone to error. This paper compares the models used and the evolution of Fintech data infrastructure developed to support financial services with …

Document-Level Relation Extraction Via Separate Relation Representation And Logical Reasoning, Heyan Huang, Changsen Yuan, Qian Liu, Yixin Cao

Research Collection School Of Computing and Information Systems

Document-level relation extraction (RE) extends the identification of entity/mentions’ relation from the single sentence to the long document. It is more realistic and poses new challenges to relation representation and reasoning skills. In this article, we propose a novel model, SRLR, using Separate Relation Representation and Logical Reasoning considering the indirect relation representation and complex reasoning of evidence sentence problems. Specifically, we first expand the judgment of relational facts from the entity-level to the mention-level, highlighting fine-grained information to capture the relation representation for the entity pair. Second, we propose a logical reasoning module to identify evidence sentences and conduct …

Hyperbolic Graph Topic Modeling Network With Continuously Updated Topic Tree, Ce Zhang, Rex Ying, Hady Wirawan Lauw

Research Collection School Of Computing and Information Systems

Connectivity across documents often exhibits a hierarchical network structure. Hyperbolic Graph Neural Networks (HGNNs) have shown promise in preserving network hierarchy. However, they do not model the notion of topics, thus document representations lack semantic interpretability. On the other hand, a corpus of documents usually has high variability in degrees of topic specificity. For example, some documents contain general content (e.g., sports), while others focus on specific themes (e.g., basketball and swimming). Topic models indeed model latent topics for semantic interpretability, but most assume a flat topic structure and ignore such semantic hierarchy. Given these two challenges, we propose a …

Semantically Constitutive Entities In Knowledge Graphs, Chong Cher Chia, Maksim Tkachenko, Hady Wirawan Lauw

Research Collection School Of Computing and Information Systems

Knowledge graphs are repositories of facts about a world. In this work, we seek to distill the set of entities or nodes in a knowledge graph into a specified number of constitutive nodes, whose embeddings would be retained. Intuitively, the remaining accessory nodes could have their original embeddings “forgotten”, and yet reconstitutable from those of the retained constitutive nodes. The constitutive nodes thus represent the semantically constitutive entities, which retain the core semantics of the knowledge graph. We propose a formulation as well as algorithmic solutions to minimize the reconstitution errors. The derived constitutive nodes are validated empirically both in …

An Adaptive Large Neighborhood Search For Heterogeneous Vehicle Routing Problem With Time Windows, Minh Pham Kien Nguyen, Aldy Gunawan, Vincent F. Yu, Mustafa Misir

Research Collection School Of Computing and Information Systems

The heterogeneous vehicle routing problem with time windows (HVRPTW) employs various vehicles with different capacities to serve upcoming pickup and delivery orders. We introduce a HVRPTW variant for reflecting the practical needs of crowd-shipping by considering the mass-rapid-transit stations, as the additional terminal points. A mixed integer linear programming model is formulated. An Adaptive Large Neighborhood Search based meta-heuristic is also developed by utilizing a basic probabilistic selection strategy, i.e. roulette wheel, and Simulated Annealing. The proposed approach is empirically evaluated on a new set of benchmark instances. The computational results revealed that ALNS shows its clear advantage on the …

Grasp Based Metaheuristic To Solve The Mixed Fleet E-Waste Collection Route Planning Problem, Aldy Gunawan, Dang V.A. Nguyen, Pham K.M. Nguyen, Pieter. Vansteenwegen

Research Collection School Of Computing and Information Systems

The digital economy has brought significant advancements in electronic devices, increasing convenience and comfort in people’s lives. However, this progress has also led to a shorter life cycle for these devices due to rapid advancements in hardware and software technology. As a result, e-waste collection and recycling have become vital for protecting the environment and people’s health. From the operations research perspective, the e-waste collection problem can be modeled as the Heterogeneous Vehicle Routing Problem with Multiple Time Windows (HVRP-MTW). This study proposes a metaheuristic based on the Greedy Randomized Adaptive Search Procedure complemented by Path Relinking (GRASP-PR) to solve …

The 4th International Workshop On Talent And Management Computing (Tmc'2023): Editorial, Hengshu Zhu, Hui Xiong, Yong Ge, Ee-Peng Lim

Research Collection School Of Computing and Information Systems

In today's competitive and fast-evolving business environment, it is a critical time for organizations to rethink how to deal with the talent and management related tasks in a quantitative manner. Indeed, thanks to the era of big data, the availability of large-scale talent data provides unparalleled opportunities for business leaders to understand the rules of talent and management, which in turn deliver intelligence for effective decision making and management for their organizations. In the past few years, talent and management computing have increasingly attracted attentions from KDD communities, and a number of research/applied data science efforts have been devoted. To …

Geometric Morphometric Analysis Of Modern Viperid Vertebrae Facilitates Identification Of Fossil Specimens, Lance D. Jessee

Electronic Theses and Dissertations

Snake vertebrae are common in the fossil record, whereas cranial remains are generally fragile and rare. Consequently, vertebrae are the most commonly studied fossil element of snakes. However, identification of snake vertebrae can be problematic due to extensive variation. This study utilizes 2-D geometric morphometrics and canonical variates analysis to 1) reveal variation between genera and species and 2) classify vertebrae of modern and fossil eastern North American Agkistrodon and Crotalus. The results show that vertebrae of Agkistrodon and Crotalus can reliably be classified to genus and species using these methods. Based on the statistical analyses, four of the …

Age And Paleontology Of The Turin Pit Locality, Monona County, Iowa, Samantha Wright

Electronic Theses and Dissertations

The Turin Pit locality (Monona county, Iowa) has been known to paleontologists since 1908, yet the age of the fauna has been unclear. Early paleontologists considered Turin Pit to date to a pre-Illinoian interglacial (the “Aftonian).” Subsequent researchers suggested it dated to the last glaciation. This study provides a partial list of mammals in the Turin Pit fauna, and together with stratigraphic information, uses the known age ranges of taxa to estimate an age for the assemblage. The presence of Mammuthus, Aenocyon, and Castoroides combined with a magnetically-reversed till located stratigraphically above fossil-bearing deposits, suggest the Turin Pit …

A Low Cost, Compact Electrochemical Analyzer Based On An Open-Source Microcontroller, Michael Kofi Darko Addo

Electronic Theses and Dissertations

Compared to other instruments for chemical analyses, electrochemical analyzers are relatively simple, inexpensive, easy to miniaturize and require little-to-no maintenance. However, like all commercially available instruments, commercial electrochemical analyzers like potentiostats primarily operate as black boxes with manufacturers providing little or no information about internal circuitry and programming. This practice can limit a researcher’s ability to develop new techniques or adapt an instrument for applications outside its typical use. In contrast, creators of open-source instruments release all the necessary information for reproduction of the hardware and software – minimizing such barriers to innovation in chemical analyses. Here, we report a …

Extraction, Purification, And Characterization Of Potential Bioactive Compounds Produced By Janthinobacterium Lividum Tajx1901, Andy Elorm Agbakpo

Electronic Theses and Dissertations

Underexplored environments such as soil samples continue to be an untapped source of bacterial strains with great potential to produce secondary metabolites for medicinal applications. As a result, these microorganisms represent a broad and yet unknown reservoir of new strains capable of producing these novel compounds. The current research primarily seeks to perform the isolation, purification, and characterization of secondary metabolites from a soil bacterium (Janthinobacterium lividum TAJX1901). The isolated soil bacterium was successfully cultured on rich media agar plates, followed by extraction using methanol and chloroform. The purification methods utilized include flash column chromatography, preparative thin-layer chromatography, and …

Isolation And Structural Determination Of Bioactive Metabolites From The Soil Bacterium, Arthrobacter Sp. Tajx1902, Amonah Arije

Electronic Theses and Dissertations

As antimicrobial resistance persistently disrupts the treatment of microbial infection, natural product chemistry has played a significant role in identifying novel drugs with novel modes of action critical to getting ahead of resistance. The primary goal of this project is to extract and identify potential novel antimicrobial natural products produced by Arthrobacter sp. Although underexplored, Arthrobacter sps. have been known to produce bioactive compounds with versatility; one such is a depsipeptide with quorum-sensing inhibitory activity.¹ In this research, an Arthrobacter sp. TAJX1902, isolated from a soil sample, showed inhibitory activity against a filamentous indicator bacterium and a violacein-producing Janthinobacterium …

The Analysis Of Extended Producer Responsibility (Epr) For E-Waste Management Policy Drivers And Challenges In Singapore, Aldy Gunawan, Tasaporn Visawameteekul, Aidan Marc Wong, Linh Chi Tran

Research Collection School Of Computing and Information Systems

This paper examines Singapore’s Extended Producer Responsibility (EPR) scheme for managing e-waste, focusing on policy drivers, challenges, and young consumers’ attitudes through an online survey. Using the Theory of Reasoned Action (TRA) and Theory of Planned Behavior (TPB), it models the relationships between awareness, norms, and perceived convenience. Findings emphasize the need for tailored policies, formal recycling promotion, and analysis of disposal factors, cost preferences, and incentives. Text analysis also highlights concerns from respondents’ recommendations.

Deep Weakly-Supervised Anomaly Detection, Guansong Pang, Chunhua Shen, Huidong Jin, Anton Van Den Hengel

Research Collection School Of Computing and Information Systems

Recent semi-supervised anomaly detection methods that are trained using small labeled anomaly examples and large unlabeled data (mostly normal data) have shown largely improved performance over unsupervised methods. However, these methods often focus on fitting abnormalities illustrated by the given anomaly examples only (i.e., seen anomalies), and consequently they fail to generalize to those that are not, i.e., new types/classes of anomaly unseen during training. To detect both seen and unseen anomalies, we introduce a novel deep weakly-supervised approach, namely Pairwise Relation prediction Network (PReNet), that learns pairwise relation features and anomaly scores by predicting the relation of any two …

Sparsity Brings Vulnerabilities: Exploring New Metrics In Backdoor Attacks, Jianwen Tian, Kefan Qiu, Debin Gao, Zhi Wang, Xiaohui Kuang, Gang Zhao

Research Collection School Of Computing and Information Systems

Nowadays, using AI-based detectors to keep pace with the fast iterating of malware has attracted a great attention. However, most AI-based malware detectors use features with vast sparse subspaces to characterize applications, which brings significant vulnerabilities to the model. To exploit this sparsityrelated vulnerability, we propose a clean-label backdoor attack consisting of a dissimilarity metric-based candidate selection and a variation ratio-based trigger construction. The proposed backdoor is verified on different datasets, including a Windows PE dataset, an Android dataset with numerical and boolean feature values, and a PDF dataset. The experimental results show that the attack can slash the accuracy …