Physical Sciences and Mathematics Commons^™

Constrained Contrastive Distribution Learning For Unsupervised Anomaly Detection And Localisation In Medical Images, Yu Tian, Guansong Pang, Fengbei Liu, Yuanhong Chen, Seon Ho Shin, Johan W. Verjans, Rajvinder Singh

Research Collection School Of Computing and Information Systems

Unsupervised anomaly detection (UAD) learns one-class classifiers exclusively with normal (i.e., healthy) images to detect any abnormal (i.e., unhealthy) samples that do not conform to the expected normal patterns. UAD has two main advantages over its fully supervised counterpart. Firstly, it is able to directly leverage large datasets available from health screening programs that contain mostly normal image samples, avoiding the costly manual labelling of abnormal samples and the subsequent issues involved in training with extremely class-imbalanced data. Further, UAD approaches can potentially detect and localise any type of lesions that deviate from the normal patterns. One significant challenge faced …

Learning To Adversarially Blur Visual Object Tracking, Qing Guo, Ziyi Cheng, Felix Juefei-Xu, Lei Ma, Xiaofei Xie, Yang Liu, Jianjun Zhao

Research Collection School Of Computing and Information Systems

Motion blur caused by the moving of the object or camera during the exposure can be a key challenge for visual object tracking, affecting tracking accuracy significantly. In this work, we explore the robustness of visual object trackers against motion blur from a new angle, i.e., adversarial blur attack (ABA). Our main objective is to online transfer input frames to their natural motion-blurred counterparts while misleading the state-of-the-art trackers during the tracking process. To this end, we first design the motion blur synthesizing method for visual tracking based on the generation principle of motion blur, considering the motion information and …

Missing Data Imputation For Solar Yield Prediction Using Temporal Multi-Modal Variational Auto-Encoder, Meng Shen, Huaizheng Zhang, Yixin Cao, Fan Yang, Yonggang Wen

Research Collection School Of Computing and Information Systems

The accurate and robust prediction of short-term solar power generation is significant for the management of modern smart grids, where solar power has become a major energy source due to its green and economical nature. However, the solar yield prediction can be difficult to conduct in the real world where hardware and network issues can make the sensors unreachable. Such data missing problem is so prevalent that it degrades the performance of deployed prediction models and even fails the model execution. In this paper, we propose a novel temporal multi-modal variational auto-encoder (TMMVAE) model, to enhance the robustness of short-term …

Causal Attention For Unbiased Visual Recognition, Tan Wang, Chang Zhou, Qianru Sun, Hanwang Zhang

Research Collection School Of Computing and Information Systems

Attention module does not always help deep models learn causal features that are robust in any confounding context, e.g., a foreground object feature is invariant to different backgrounds. This is because the confounders trick the attention to capture spurious correlations that benefit the prediction when the training and testing data are IID (identical & independent distribution); while harm the prediction when the data are OOD (out-of-distribution). The sole fundamental solution to learn causal attention is by causal intervention, which requires additional annotations of the confounders, e.g., a “dog” model is learned within “grass+dog” and “road+dog” respectively, so the “grass” and …

Disentangling Hate In Online Memes, Ka Wei, Roy Lee, Rui Cao, Ziqing Fan, Jing Jiang, Wen Haw Chong

Research Collection School Of Computing and Information Systems

Hateful and offensive content detection has been extensively explored in a single modality such as text. However, such toxic information could also be communicated via multimodal content such as online memes. Therefore, detecting multimodal hateful content has recently garnered much attention in academic and industry research communities. This paper aims to contribute to this emerging research topic by proposing DisMultiHate, which is a novel framework that performed the classification of multimodal hateful content. Specifically, DisMultiHate is designed to disentangle target entities in multimodal memes to improve the hateful content classification and explainability. We conduct extensive experiments on two publicly available …

Measuring Data Collection Diligence For Community Healthcare, Galawala Ramesha Samurdhi Karunasena, M. S. Ambiya, Arunesh Sinha, R. Nagar, S. Dalal, Abdullah. H., D. Thakkar, D. Narayanan, M. Tambe

Research Collection School Of Computing and Information Systems

Data analytics has tremendous potential to provide targeted benefit in low-resource communities, however the availability of highquality public health data is a significant challenge in developing countries primarily due to non-diligent data collection by community health workers (CHWs). Our use of the word non-diligence here is to emphasize that poor data collection is often not a deliberate action by CHW but arises due to a myriad of factors, sometime beyond the control of the CHW. In this work, we define and test a data collection diligence score. This challenging unlabeled data problem is handled by building upon domain expert’s guidance …

Multi-Modal Recommender Systems: Hands-On Exploration, Quoc Tuan Truong, Aghiles Salah, Hady Wirawan Lauw

Research Collection School Of Computing and Information Systems

Recommender systems typically learn from user-item preference data such as ratings and clicks. This information is sparse in nature, i.e., observed user-item preferences often represent less than 5% of possible interactions. One promising direction to alleviate data sparsity is to leverage auxiliary information that may encode additional clues on how users consume items. Examples of such data (referred to as modalities) are social networks, item’s descriptive text, product images. The objective of this tutorial is to offer a comprehensive review of recent advances to represent, transform and incorporate the different modalities into recommendation models. Moreover, through practical hands-on sessions, we …

On The Usability (In)Security Of In-App Browsing Interfaces In Mobile Apps, Zicheng Zhang, Daoyuan Wu, Lixiang Li, Debin Gao

Research Collection School Of Computing and Information Systems

Due to the frequent encountering of web URLs in various application scenarios (e.g., chatting and email reading), many mobile apps build their in-app browsing interfaces (IABIs) to provide a seamless user experience. Although this achieves user-friendliness by avoiding the constant switching between the subject app and the system built-in browser apps, we find that IABIs, if not well designed or customized, could result in usability security risks. In this paper, we conduct the first empirical study on the usability (in)security of in-app browsing interfaces in both Android and iOS apps. Specifically, we collect a dataset of 25 high-profile mobile apps …

Integrated Discourse Analysis & Learning Skills Framework For Class Conversations, Devyn Wei Hung Tan, Gottipati Swapna, Kyong Jin Shim, Shankararaman, Venky

Research Collection School Of Computing and Information Systems

Constructive interactions through discussion forums allow students to open their horizons and thought processes to acquire more knowledge and develop skills. Thus, discussion forums play an important role in supporting learning. Additionally, the discussion forum provides the content for creating a knowledge repository. It contains discussion threads related to key course topics that are debated by the students. One approach to understanding the student learning experience is through the analysis of the discussion threads. This research proposes the application of discourse analysis and collaborative learning frameworks to discussion forums to gain further insights into the student’s learning in a classroom. …

Latent Class Analysis For Identifying Subclasses Of Depression Using Jmp Pro 16, Karishma Yadav, Fei Fei Sue-Ann Seet, Tin Seong Kam, Tin Seong Kam

Research Collection School Of Computing and Information Systems

According to WHO, “Depression is a leading cause of disability worldwide and is a major contributor to the overall global burden of disease”. A major stumbling block in the care of depressed patients remains the accurate diagnosis of the severity of depression. Patient Health Questionnaire (PHQ-9), a 9-question instrument is widely used for diagnosing and determining the severity of depression. However, the popularly used 5-Category of depression severity based on the sum of responses to the 9 questions was overly subjective. In view of this limitation, our paper aims to demonstrate how Latent Class Analysis of JMP Pro can be …

Transporting Causal Mechanisms For Unsupervised Domain Adaptation, Zhongqi Yue, Qianru Sun, Xian-Sheng Hua, Hanwang Zhang

Research Collection School Of Computing and Information Systems

Existing Unsupervised Domain Adaptation (UDA) literature adopts the covariate shift and conditional shift assumptions, which essentially encourage models to learn common features across domains. However, due to the lack of supervision in the target domain, they suffer from the semantic loss: the feature will inevitably lose nondiscriminative semantics in source domain, which is however discriminative in target domain. We use a causal view—transportability theory [41]—to identify that such loss is in fact a confounding effect, which can only be removed by causal intervention. However, the theoretical solution provided by transportability is far from practical for UDA, because it requires the …

Prediction Of Synthetic Lethal Interactions In Human Cancers Using Multi-View Graph Auto-Encoder, Zhifeng Hao, Di Wu, Yuan Fang, Min Wu, Ruichu Cai, Xiaoli Li

Research Collection School Of Computing and Information Systems

Synthetic lethality (SL) is a very important concept for the development of targeted anticancer drugs. However, experimental methods for SL detection often suffer from various issues like high cost and low consistency across cell lines. Hence, computational methods for predicting novel SLs have recently emerged as complements for wet-lab experiments. In addition, SL data can be represented as a graph where nodes are genes and edges are the SL interactions. It is thus motivated to design advanced graph-based machine learning algorithms for SL prediction. In this paper, we propose a novel SL prediction method using Multi-view Graph Auto-Encoder (SLMGAE). We …

Cloud, Edge And Fog Computing: Trends And Case Studies, Eng Lieh Ouh, Stanislaw Jarzabek, Geok Shan Lim, Masayoshi Ogawa

Research Collection School Of Computing and Information Systems

As it is done today, an informal – solely based on experts’ intuition – evaluation of profitability of adopting cloud services is undependable and not scalable as there are many conflicting factors and constraints such evaluation should account for. The revenue from service tenants and the cost of implementing the service architecture are the leading service factors that drive profitability. Cloud service architectures also need to handle a growing number of tenants with increasingly diverse requirements which must be weighed against the capabilities and costs of various service architectures, particularly single- versus multi-tenanted models. We believe a conceptual model enumerating …

Assessing Generalizability Of Codebert, Xin Zhou, Donggyun Han, David Lo

Research Collection School Of Computing and Information Systems

Pre-trained models like BERT have achieved strong improvements on many natural language processing (NLP) tasks, showing their great generalizability. The success of pre-trained models in NLP inspires pre-trained models for programming language. Recently, CodeBERT, a model for both natural language (NL) and programming language (PL), pre-trained on code search dataset, is proposed. Although promising, CodeBERT has not been evaluated beyond its pre-trained dataset for NL-PL tasks. Also, it has only been shown effective on two tasks that are close in nature to its pre-trained data. This raises two questions: Can CodeBERT generalize beyond its pre-trained data? Can it generalize to …

Condensing A Sequence To One Informative Frame For Video Recognition, Qiu. Zhaofan, Ting Yao, Yan Shu, Chong-Wah Ngo, Tao Mei

Research Collection School Of Computing and Information Systems

Video is complex due to large variations in motion and rich content in fine-grained visual details. Abstracting useful information from such information-intensive media requires exhaustive computing resources. This paper studies a two-step alternative that first condenses the video sequence to an informative" frame" and then exploits off-the-shelf image recognition system on the synthetic frame. A valid question is how to define" useful information" and then distill it from a video sequence down to one synthetic frame. This paper presents a novel Informative Frame Synthesis (IFS) architecture that incorporates three objective tasks, ie, appearance reconstruction, video categorization, motion estimation, and two …

Land Use As A Predictor Of Water Hyacinth (Eichhornia Crassipes) Presence On The Entebbe Coast Of Lake Victoria, Uganda, Zachary Hoffman

Independent Study Project (ISP) Collection

Lake Victoria is shared amongst Tanzania, Kenya, and Uganda and has tremendous ecological, economical, and cultural significance. Within the lake system, there are several problems, including the proliferation of an invasive weed, water hyacinth (Eichhornia crassipes). Therefore, this study aims to assess several factors that may correlate with water hyacinth proliferation. The specific objectives are (1) to identify possible correlations of water hyacinth density and land use around Entebbe, Uganda, and (2) to identify annual trends in water hyacinth coverage, to better inform policy and conservation efforts. Entebbe has a coastline of six land cover types: flooded vegetation, trees, grasses, …

Geology And Hydrogeology Of Northeastern Nebraska: Geology, Water Management And Geological Hazards, Nebraska Geological Society Field Trip 2021, Robert F. Diffendal Jr., Sue Olafsen Lackey, Douglas R. Hallum

Conservation and Survey Division

The 2021 Nebraska Geological Society field trip will provide an overview of the geology and hydrogeology of northeast Nebraska, the groundwater quality and quantity issues unique to the region, water management issues and decision-making, and ongoing work relating to water quality and geologic (hydrostratigraphic) framework, along the lower Niobrara and Missouri rivers Between Spencer Dam and Homer Nebraska. This program will be comprised of an auto tour with multiple stops on Saturday October 16 through Monday October 18 as illustrated in Map 1. The tour will begin with an examination of the Spencer Dam site and several locations in and …

Timescales And Dynamics Of Rhyolitic Melt Generation In Upper Crustal Magmatic Reservoirs: Insights From New Geochronologic And Petrologic Data On The Valle Mosso Pluton And Sesia Caldera, Sesia Magmatic System, Italy, Lorenzo Tavazzani

Earth Sciences Theses and Dissertations

This work may have a narrow geographic focus, namely a handful of steep and densely forested valleys of the Southern Italian Alps, but its scientific goals are certainly wide. Through a combined analyses of the spatial distribution, geochemical characteristics and temporal duration of the Permian silicic magmatism in the renown Ivrea crustal section (Sesia Magmatic System), this dissertation addresses fundamental questions regarding the physical mechanisms of rhyolitic magma storage, mobilization and eruption.

A brief introduction serves to define the broader context of the scientific questions addressed in the three main chapters of this dissertation. Through Chapters 2 to 4 we …

Beach Suitability For Nesting Olive Ridley Sea Turtles (Lepidochelys Olivacea), Playa Malena, Azuero Peninsula, Panama, Soleil Foy

Independent Study Project (ISP) Collection

Sea turtle species all over the world are facing continuous declines in their populations partly due to diminished viable nesting habitat. Understanding sea turtle nesting preferences and the suitability of beaches for reproductive success is necessary in order to sustain their populations and ensure their conservation. Olive ridley sea turtles (Lepidochelys olivacea) are one of the species experiencing this trend, especially in the Eastern Pacific. This study examined the beach suitability of a site known as Playa Malena in the Azuero Peninsula in Panama for Olive ridley nesting. This site was compared to the suitability of 7 other …

A Ladder Of Endemicity: A Pioneering Study Of Anuran Communities Along An Elevational Gradient In The Eastern Andean Cloud Forests, Ecuador, Ella Halbert

Independent Study Project (ISP) Collection

Per unit of area, Ecuador is home to the most herpetofauna of any region on Earth, yet many ecosystems are under researched. Positioned between the Eastern Andes and the Amazon rainforest, the Río Pastaza Basin of Ecuador has been the recent focus of anuran biodiversity surveys and is home to many endemic species. This study aimed to determine how the anuran community composition of a vertical transect in the Llanganates National Park varied between elevations. A 7-day expedition was conducted along a new trail up Abitagua Range spanning 1,050 m in elevation. Each night, visual encounter surveys were employed to …

Political Processes Of Displacement In Infrastructure Development: The Case Of Aldeia Da Luz And The Alqueva Dam, Cordelia Walz

Independent Study Project (ISP) Collection

In 2002, the residents of Aldeia da Luz left their village for the final time, displaced to make room for the Alqueva Dam, a massive regional hydropower and irrigation project almost half a century in the making. The Alqueva Multipurpose Project was marketed by the Portuguese government as a way to develop the impoverished region of the Alentejo and bring innovation to the agricultural sector. The village of Luz was the only physical obstacle to this goal and, therefore, its residents were dispossessed, sacrificed for the development of the greater Alentejo region. However, unlike many other large-scale infrastructure projects that …

Taking The Waste Out Of Wastewater: Evaluation And Implementation Of Sustainability Measures At Base Aérea N.°5’S Wastewater Treatment Plant, Abigail Monahan

Independent Study Project (ISP) Collection

Wastewater treatment is a critical process by which the concentration of pollutants in wastewater are reduced to minimize the resulting effluent’s impact on human health and the environment. Base Aérea N.°5, an airbase located in Monte Real, Portugal, is striving to enhance their wastewater treatment methods to decrease the impact of the treated effluent on the environment. They can reduce the effluent’s impact by implementing more advanced wastewater treatment to decrease the concentration of micropollutants and pathogens in the effluent, and they can reuse the treated effluent to decrease the volume of freshwater they extract. This work seeks to evaluate …

Public Perception Of Forest And Fire Management Policy In Portugal, Jensen Rocha

Independent Study Project (ISP) Collection

The impacts of forest fires are growing as climate change creates more frequent ideal temperature conditions for extreme wildfires. With the devastating fires it experienced in 2017, Portugal stands as an example of the impacts of dangerous wildfires and is working on developing a plan to protect its country from this rising threat. Important in the plan development is the opinion of the many stakeholders involved and the public whom the fires impact. Using an interview and survey approach, this study gathers public and stakeholder opinion on the forest and fire management policy in Portugal to understand what issues in …

The Role Of Public And Private Sectors: How To Promote National Cybersecurity Strategies And Critical Infrastructure Protection In Southeastern Europe, Larry Cruz

Independent Study Project (ISP) Collection

This paper seeks to address the important role of public and private interests in protecting critical infrastructure in Southeastern Europe, providing examples from Serbia. While the public sector does have a role in protecting critical infrastructure needs, it is the private sector which holds major oversight of the critical infrastructures of the region, therefore having an important role in maintaining their functionality and protection. The literature in this field argues for more collaboration and information sharing between the public and private sectors of the region, though the task is not as simple as it appears given the varying aspirations of …

The Labyrinth Of Data Collection For Humanitarian Project Funding And Implementation, Maria Alejandra Pulido

Independent Study Project (ISP) Collection

My research concentrates on four NGOs: IOM, IDMC, JIPS, and OCHA which use different tools to collect data and translate the information into evidence for data-driven decision making (DDDM) for the implementation of humanitarian assistance projects. I focus on the importance, advantages, and various data collection tools which help ameliorate the humanitarian sector since it does not have a current professionalized path to enter the workforce. I incorporated four interviews, attended two conferences and analyzed multiple online sources during my project.

Investigating The Impact Of Land Use On Avian Diversity And Abundance In Areas Surrounding Mabamba Swamp, Uganda, Sydney Marie Jones

Independent Study Project (ISP) Collection

The primary objective of this study was to investigate the impact of wetland-adjacent land use on avian species richness and abundance areas surrounding Mabamba Swamp, Uganda. Four types of land use were investigated: Eucalyptus plantations, wetland-edge agricultural fields, residential areas, and mature secondary forests. A total of 40-morning point counts were conducted for ten days in late November and late December of 2021. One-way ANOVA tests and Tukey’s HSD tests revealed significant differences in mean avian richness and abundance between all sites except residential areas and Nkima Forest. Additionally, Nkima Forest was found to contain the most number of specialist …

Man V.S. Wild: An Analysis Of Language Used Regarding Human-Wildlife Conflict In The Kibale National Park Community, Western Region, Uganda, Sophie Perfetto

Independent Study Project (ISP) Collection

If humans are to live sustainably and in harmony with wildlife in the vicinity of Protected Areas, policy makers, and government authorities need to make informed decisions with consideration to the needs of local communities. For this to happen, policy makers must understand the perceptions of local communities and take local perspectives into account. As language is at the root of perception, language was studied in the context of its role in shaping local perceptions of human-wildlife conflict and consequential conflict mitigation strategies. Six communities around the Northern region of Kibale National Park, Uganda, were studied, with sites located in …

Using Concurrent Functional Regression To Reconstruct River Stage Data During Flood Events And Identify Influential Functional Measurements, Ryan Pittman

Theses and Dissertations

On October 4, 2015, the Cedar Creek gage at Congaree National Park stopped reporting stages, and the readings did not resume until approximately two weeks later because of record-breaking rainfall that led to some of the worst flooding in South Carolina history. Our goal is to reconstruct the Cedar Creek stage during this missing two-week window. The Congaree River gage in Congaree National Park remained functioning throughout the October 2015 flood, when the stage reached its maximum recorded crest. The stages from the two gages are directly related during floods as water travels through the local spillways and flood planes …

Correct Web Service Transactions In The Presence Of Malicious And Misbehaving Transactions, John Thomas Ravan Iii

Theses and Dissertations

Concurrent database transactions within a web service environment can cause a variety of problems without the proper concurrency control mechanisms in place. A few of these problems involve data integrity issues, deadlock, and efficiency issues. Even with today’s industry standard solutions to these problems, they have taken a reactive approach rather than proactively preventing these problems from happening. We deliver a solution, based on prediction-based scheduling to ensure consistency while keeping execution time the same or faster than current industry solutions. The first part of this solution involves prototyping and formally proving a prediction-based scheduler.

The prediction-based scheduler leverages a …

Multi-Objective Routing For Distributed Controllers, Konstantin Y. Rubin

Theses and Dissertations

A long-term goal of future naval shipboard power systems is the ability to manage energy flow with sufficient flexibility to accommodate future platform requirements such as better survivability, continuity, and support of pulsed and other demanding loads. To facilitate scalable, low-latency global distributed system control, each control module can include an integrated network interface connected through multiple channels onto a direct, multi-hop network topology. In this work, we focus on a 2D Torus, in which control nodes are arranged in a regular 2D grid, with each node connected through point-to-point connections to its four immediate neighbors. An important advantage of …