Engineering Commons^™

Solid Spherical Energy (Sse) Cnns For Efficient 3d Medical Image Analysis, Vincent Andrearczyk, Valentin Oreiller, Julien Fageot, Xavier Montet, Adrien Depeursinge

Session 2: Deep Learning for Computer Vision

Invariance to local rotation, to differentiate from the global rotation of images and objects, is required in various texture analysis problems. It has led to several breakthrough methods such as local binary patterns, maximum response and steerable filterbanks. In particular, textures in medical images often exhibit local structures at arbitrary orientations. Locally Rotation Invariant (LRI) Convolutional Neural Networks (CNN) were recently proposed using 3D steerable filters to combine LRI with Directional Sensitivity (DS). The steerability avoids the expensive cost of convolutions with rotated kernels and comes with a parametric representation that results in a drastic reduction of the number of …

Adapted K-Nearest Neighbors For Detecting Anomalies On Spatio–Temporal Traffic Flow, Youcef Djenouri, Asma Belhadi, Jerry Chun-Wei Lin, Alberto Cano

Computer Science Publications

Outlier detection is an extensive research area, which has been intensively studied in several domains such as biological sciences, medical diagnosis, surveillance, and traffic anomaly detection. This paper explores advances in the outlier detection area by finding anomalies in spatio-temporal urban traffic flow. It proposes a new approach by considering the distribution of the flows in a given time interval. The flow distribution probability (FDP) databases are first constructed from the traffic flows by considering both spatial and temporal information. The outlier detection mechanism is then applied to the coming flow distribution probabilities, the inliers are stored to enrich the …

An Investigation Of Three Subjective Rating Scales Of Mental Workload In Third Level Education, Nha Vu Thanh Nguyen

Dissertations

Mental Workload assessment in educational settings is still recognized as an open research problem. Although its application is useful for instructional design, it is still unclear how it can be formally shaped and which factors compose it. This paper is aimed at investigating a set of features believed to shape the construct of mental workload and aggregating them together in models trained with supervised machine learning techniques. In detail, multiple linear regression and decision trees have been chosen for training models with features extracted respectively from the NASA Task Load Index and the Workload Profile, well-known self-reporting instruments for assessing …

Forecasting Anomalous Events And Performance Correlation Analysis In Event Data, Sonya Leech [Thesis]

Dissertations

Classical and Deep Learning methods are quite common approaches for anomaly detection. Extensive research has been conducted on single point anomalies. Collective anomalies that occur over a set of two or more durations are less likely to happen by chance than that of a single point anomaly. Being able to observe and predict these anomalous events may reduce the risk of a server’s performance. This paper presents a comparative analysis into time-series forecasting of collective anomalous events using two procedures. One is a classical SARIMA model and the other is a deep learning Long-Short Term Memory (LSTM) model. It then …

An Investigation Into The Predictive Capability Of Customer Spending In Modelling Mortgage Default, Donal Finn [Thesis]

Dissertations

The mortgage arrears crisis in Ireland was and is among the most severe experienced on record and although there has been a decreasing trend in the number of mortgages in default in the past four years, it still continues to cause distress to borrowers and vulnerabilities to lenders. There are indications that one of the main factors associated with mortgage default is loan affordability, of which the level of disposable income is a driver. Additionally, guidelines set out by the European Central Bank instructed financial institutions to adopt measures to further reduce and prevent loans defaulting, including the implementation and …

Human Action Recognition In Videos Using Transfer Learning, Kaiqiang Huang, Sarah Jane Delany, Susan Mckeever

Session 1: Active Vision, Tracking, Motion Analysis

A variety of systems focus on detecting the actions and activities performed by humans, such as video surveillance and health monitoring systems. However, published labelled human action datasets for training supervised machine learning models are limited in number and expensive to produce. The use of transfer learning for the task of action recognition can help to address this issue by transferring or re-using the knowledge of existing trained models, in combination with minimal training data from the new target domain. Our focus in this paper is an investigation of video feature representations and machine learning algorithms for transfer learning for …

Micro Expression Classification Accuracy Assessment, Pratikshya Sharma, Sonya Coleman, Pratheepan Yogarajah, Laurenc Taggart

Session 1: Active Vision, Tracking, Motion Analysis

The ability to identify and draw appropriate implications from non-verbal cues is a challenging task in facial expression recognition and has been investigated by various disciplines particularly social science, medical science, psychology and technological sciences beyond three decades. Non-verbal cues often last a few seconds and are obvious (macro) whereas others are very short and difficult to interpret (micro). This research is based on the area of micro expression recognition with the main focus laid on understanding and exploring the combined effect of various existing feature extraction techniques and one of the most renowned machine learning algorithms identified as Support …

Utilizing Deep Neural Networks For Brain–Computer Interface-Based Prosthesis Control, Thomas C. Noel, Brian R. Snider

Faculty Publications - Department of Electrical Engineering and Computer Science

Limb amputations affect a significant portion of the world’s population every year. The necessity for these operations can be associated with related health conditions or a traumatic event. Currently, prosthetic devices intended to alleviate the burden of amputation lack many of the premier features possessed by their biological counterparts. The foremost of these features are agility and tactile function. In an effort to address the former, researchers here investigate the fundamental connection between agile finger movement and brain signaling. In this study each subject was asked to move his or her right index finger in sync with a time-aligned finger …

Toward The Personalization Of Copyright Law, Adi Libson, Gideon Parchomovsky

All Faculty Scholarship

In this Article, we provide a blueprint for personalizing copyright law in order to reduce the deadweight loss that stems from its universal application to all users, including those who would not have paid for it. We demonstrate how big data can help identify inframarginal users, who would not pay for copyrighted content, and we explain how copyright liability and remedies should be modified in such cases.

Deep Learning: Edge-Cloud Data Analytics For Iot, Katarina Grolinger, Ananda M. Ghosh

Electrical and Computer Engineering Publications

Sensors, wearables, mobile and other Internet of Thing (IoT) devices are becoming increasingly integrated in all aspects of our lives. They are capable of collecting massive quantities of data that are typically transmitted to the cloud for processing. However, this results in increased network traffic and latencies. Edge computing has a potential to remedy these challenges by moving computation physically closer to the network edge where data are generated. However, edge computing does not have sufficient resources for complex data analytics tasks. Consequently, this paper investigates merging cloud and edge computing for IoT data analytics and presents a deep learning-based …

Machine Learning And Neural Networks For Real-Time Scheduling, Daniel Hureira, Christian Vartanian

Recent Advances in Real-Time Systems

This paper aims to serve as an efficient survey of the processes, problems, and methodologies surrounding the use of Neural Networks, specifically Hopfield-Type, in order to solve Hard-Real-Time Scheduling problems. Our primary goal is to demystify the field of Neural Networks research and properly describe the methods in which Real-Time scheduling problems may be approached when using neural networks. Furthermore, to give an introduction of sorts on this niche topic in a niche field. This survey is derived from four main papers, namely: “A Neurodynamic Approach for Real-Time Scheduling via Maximizing Piecewise Linear Utility” and “Scheduling Multiprocessor Job with Resource …

Distance,Time And Terms In First Story Detection, Fei Wang

Doctoral

First Story Detection (FSD) is an important application of online novelty detection within Natural Language Processing (NLP). Given a stream of documents, or stories, about news events in a chronological order, the goal of FSD is to identify the very first story for each event. While a variety of NLP techniques have been applied to the task, FSD remains challenging because it is still not clear what is the most crucial factor in defining the “story novelty”. Giventhesechallenges,thethesisaddressedinthisdissertationisthat the notion of novelty in FSD is multi-dimensional. To address this, the work presented has adopted a three dimensional analysis of the …

Music Information Retrieval For Irish Traditional Music Automatic Analysis Of Harmonic, Rhythmic, And Melodic Features For Efficient Key-Invariant Tune Recognition, Pierre Beauguitte

Doctoral

Music making and listening practices increasingly rely on techno logy,and,asaconsequence,techniquesdevelopedinmusicinformation retrieval (MIR) research are more readily available to end users, in par ticular via online tools and smartphone apps. However, the majority of MIRresearchfocusesonWesternpopandclassicalmusic,andthusdoes not address specificities of other musical idioms. Irishtraditionalmusic(ITM)ispopularacrosstheglobe,withregular sessionsorganisedonallcontinents. ITMisadistinctivemusicalidiom, particularly in terms of heterophony and modality, and these character istics can constitute challenges for existing MIR algorithms. The bene fitsofdevelopingMIRmethodsspecificallytailoredtoITMisevidenced by Tunepal, a query-by-playing tool that has become popular among ITM practitioners since its release in 2009. As of today, Tunepal is the state of the art for tune recognition in ITM. The research in …

An Explainable Autoencoder For Collaborative Filtering Recommendation, Pegah Sagheb Haghighi, Olurotimi Seton, Olfa Nasraoui

Faculty Scholarship

Autoencoders are a common building block of Deep Learning architectures, where they are mainly used for representation learning. They have also been successfully used in Collaborative Filtering (CF) recommender systems to predict missing ratings. Unfortunately, like all black box machine learning models, they are unable to explain their outputs. Hence, while predictions from an Autoencoderbased recommender system might be accurate, it might not be clear to the user why a recommendation was generated. In this work, we design an explainable recommendation system using an Autoencoder model whose predictions can be explained using the neighborhood based explanation style. Our preliminary work …

Big Data Investment And Knowledge Integration In Academic Libraries, Saher Manaseer, Afnan R. Alawneh, Dua Asoudi

Copyright, Fair Use, Scholarly Communication, etc.

Recently, big data investment has become important for organizations, especially with the fast growth of data following the huge expansion in the usage of social media applications, and websites. Many organizations depend on extracting and reaching the needed reports and statistics. As the investments on big data and its storage have become major challenges for organizations, many technologies and methods have been developed to tackle those challenges.

One of such technologies is Hadoop, a framework that is used to divide big data into packages and distribute those packages through nodes to be processed, consuming less cost than the traditional storage …

Predicting Customer Retention Of An App-Based Business Using Supervised Machine Learning, Jeswin Jose

Dissertations

Identification of retainable customers is very essential for the functioning and growth of any business. An effective identification of retainable customers can help the business to identify the reasons of retention and plan their marketing strategies accordingly. This research is aimed at developing a machine learning model that can precisely predict the retainable customers from the total customer data of an e-learning business. Building predictive models that can efficiently classify imbalanced data is a major challenge in data mining and machine learning. Most of the machine learning algorithms deliver a suboptimal performance when introduced to an imbalanced dataset. A variety …

A Graph-Based Reinforcement Learning Method With Converged State Exploration And Exploitation, Han Li, Tianding Chen, Hualiang Teng, Yingtao Jiang

Civil and Environmental Engineering and Construction Faculty Research

In any classical value-based reinforcement learning method, an agent, despite of its continuous interactions with the environment, is yet unable to quickly generate a complete and independent description of the entire environment, leaving the learning method to struggle with a difficult dilemma of choosing between the two tasks, namely exploration and exploitation. This problem becomes more pronounced when the agent has to deal with a dynamic environment, of which the configuration and/or parameters are constantly changing. In this paper, this problem is approached by first mapping a reinforcement learning scheme to a directed graph, and the set that contains all …

Sdn Testbed For Evaluation Of Large Exo-Atmospheric Emp Attacks, Diogo Oliveira, Nasir Ghani, Majeed M. Hayat, Jorge Crichigno, Elias Bou-Harb

Electrical and Computer Engineering Faculty Research and Publications

Large-scale nuclear electromagnetic pulse (EMP) attacks and natural disasters can cause extensive network failures across wide geographic regions. Although operational networks are designed to handle most single or dual faults, recent efforts have also focused on more capable multi-failure disaster recovery schemes. Concurrently, advances in software-defined networking (SDN) technologies have delivered highly-adaptable frameworks for implementing new and improved service provisioning and recovery paradigms in real-world settings. Hence this study leverages these new innovations to develop a robust disaster recovery (counter-EMP) framework for large backbone networks. Detailed findings from an experimental testbed study are also presented.

Vireo @ Video Browser Showdown 2019, Phuong Anh Nguyen, Chong-Wah Ngo, Danny Francis, Benoit Huet

Research Collection School Of Computing and Information Systems

In this paper, the VIREO team video retrieval tool is described in details. As learned from Video Browser Showdown (VBS) 2018, the visualization of video frames is a critical need to improve the browsing effectiveness. Based on this observation, a hierarchical structure that represents the video frame clusters has been built automatically using k-means and self-organizing-map and used for visualization. Also, the relevance feedback module which relies on real-time supportvector-machine classification becomes unfeasible with the large dataset provided in VBS 2019 and has been replaced by a browsing module with pre-calculated nearest neighbors. The preliminary user study results on IACC.3 …

Shaping The Future Of Trusted Digital Identity: The Idef Registry With Health Information Use Cases, Noreen Y. Whysel

Publications and Research

No abstract provided.

Computational Modeling Of Trust Factors Using Reinforcement Learning, C. M. Kuzio, A. Dinh, C. Stone, L. Vidyaratne, K. M. Iftekharuddin

Electrical & Computer Engineering Faculty Publications

As machine-learning algorithms continue to expand their scope and approach more ambiguous goals, they may be required to make decisions based on data that is often incomplete, imprecise, and uncertain. The capabilities of these models must, in turn, evolve to meet the increasingly complex challenges associated with the deployment and integration of intelligent systems into modern society. Historical variability in the performance of traditional machine-learning models in dynamic environments leads to ambiguity of trust in decisions made by such algorithms. Consequently, the objective of this work is to develop a novel computational model that effectively quantifies the reliability of autonomous …

Microcontroller Based Granular Urea Application Attachment For Rice Transplanter, Md Towfiqur Rahman, Md Monjurul Alam, Md Mosharraf Hossain, Muhammad Rashed Al Mamun

Biological Systems Engineering: Papers and Publications

Transplanting and fertilizer application for rice production in Bangladesh are tedious, time consuming and laborious task, and mostly done manually. Mechanical transplanting of rice becoming popular in the country in recent years and few machines have been developed for granular urea deep placement, however, having some limitations. Placing granular urea precisely along with rice transplanting, an attempt was under taken to design and fabricate an electronic control granular urea applicator to be attach with a 4-row walk behind type rice transplanter. Fabrication of the electronic granular urea applicator was done in the workshop of the Department of Farm Power and …

Identifying Key Topics Bearing Negative Sentiment On Twitter: Insights Concerning The 2015-2016 Zika Epidemic, Ravali Mamidi, Michele Miller, Tanvi Banerjee, William Romine, Amit Sheth

Publications

Background To understand the public sentiment regarding the Zika virus, social media can be leveraged to understand how positive, negative, and neutral sentiments are expressed in society. Specifically, understanding the characteristics of negative sentiment could help inform federal disease control agencies’ efforts to disseminate relevant information to the public about Zika-related issues.

Objective The purpose of this study was to analyze the public sentiment concerning Zika using posts on Twitter and determine the qualitative characteristics of positive, negative, and neutral sentiments expressed.

Methods Machine learning techniques and algorithms were used to analyze the sentiment of tweets concerning Zika. A supervised …

Jitana: A Modern Hybrid Program Analysis Framework For Android Platforms, Yutaka Tsutano, Shakthi Bachala, Witawas Srisa-An, Gregg Rothermel, Jackson Dinh

School of Computing: Faculty Publications

Security vetting of Android apps is often performed under tight time constraints (e.g., a few minutes). As such, vetting activities must be performed “at speed”, when an app is submitted for distribution or a device is analyzed for malware. Existing static and dynamic program analysis approaches are not feasible for use in security analysis tools because they require a much longer time to operate than security analysts can afford. There are two factors that limit the performance and efficiency of current analysis approaches. First, existing approaches analyze only one app at a time. Finding security vulnerabilities in collaborative environments such …

[Accepted Article Manuscript Version (Postprint)] Distributed Data-Gathering And -Processing In Smart Cities: An Information-Centric Approach, Reza Tourani, Abderrahmen Mtibaa, Satyajayant Misra

Computer Science Faculty Works

The technological advancements along with the proliferation of smart and connected devices (things) motivated the exploration of the creation of smart cities aimed at improving the quality of life, economic growth, and efficient resource utilization. Some recent initiatives defined a smart city network as the interconnection of the existing independent and heterogeneous networks and the infrastructure. However, considering the heterogeneity of the devices, communication technologies, network protocols, and platforms the interoperability of these networks is a challenge requiring more attention. In this paper, we propose the design of a novel Information-Centric Smart City architecture (iSmart), focusing on the demand of …

Determination Of Personalized Asthma Triggers From Multimodal Sensing And A Mobile App: Observational Study, Revathy Venkataramanan, Krishnaprasad Thirunarayan, Utkarshani Jaimini, Dipesh Kadariya, Hong Yung Yip, Maninder Kalra, Amit Sheth

Publications

Background: Asthma is a chronic pulmonary disease with multiple triggers. It can be managed by strict adherence to an asthma care plan and by avoiding these triggers. Clinicians cannot continuously monitor their patients’ environment and their adherence to an asthma care plan, which poses a significant challenge for asthma management.

Objective: In this study, pediatric patients were continuously monitored using low-cost sensors to collect asthma-relevant information. The objective of this study was to assess whether kHealth kit, which contains low-cost sensors, can identify personalized triggers and provide actionable insights to clinicians for the development of a tailored asthma care plan. …

Predicting Cascading Failures In Power Grids Using Machine Learning Algorithms, Rezoan Ahmed Shuvro, Pankaz Das, Majeed M. Hayat, Mitun Talukder

Electrical and Computer Engineering Faculty Research and Publications

Although there has been notable progress in modeling cascading failures in power grids, few works included using machine learning algorithms. In this paper, cascading failures that lead to massive blackouts in power grids are predicted and classified into no, small, and large cascades using machine learning algorithms. Cascading-failure data is generated using a cascading failure simulator framework developed earlier. The data set includes the power grid operating parameters such as loading level, level of load shedding, the capacity of the failed lines, and the topological parameters such as edge betweenness centrality and the average shortest distance for numerous combinations of …

Internet Of Things In Smart Agriculture: Enabling Technologies, Abdul Salam, Syed Shah

Faculty Publications

In this paper, an IoT technology research and innovation roadmap for the field of precision agriculture (PA) is presented. Many recent practical trends and the challenges have been highlighted. Some important objectives for integrated technology research and education in precision agriculture are described. Effective IoT based communications and sensing approaches to mitigate challenges in the area of precision agriculture are presented.

Inception: Virtual Space In Memory Space In Real Space, Peter Casey, Rebecca Lindsay-Decusati, Ibrahim Baggili, Frank Breitinger

Electrical & Computer Engineering and Computer Science Faculty Publications

Virtual Reality (VR) has become a reality. With the technology's increased use cases, comes its misuse. Malware affecting the Virtual Environment (VE) may prevent an investigator from ascertaining virtual information from a physical scene, or from traditional “dead” analysis. Following the trend of antiforensics, evidence of an attack may only be found in memory, along with many other volatile data points. Our work provides the primary account for the memory forensics of Immersive VR systems, and in specific the HTC Vive. Our approach is capable of reconstituting artifacts from memory that are relevant to the VE, and is also capable …

Real-Time Classification Of Multivariate Olfaction Data Using Spiking Neural Networks, Arnup Vanarse, Adam Osseiran, Alexander Rassau, Therese O'Sullivan, Jonny Lo, Amanda Devine

Research outputs 2014 to 2021

Recent studies in bioinspired artificial olfaction, especially those detailing the application of spike-based neuromorphic methods, have led to promising developments towards overcoming the limitations of traditional approaches, such as complexity in handling multivariate data, computational and power requirements, poor accuracy, and substantial delay for processing and classification of odors. Rank-order-based olfactory systems provide an interesting approach for detection of target gases by encoding multi-variate data generated by artificial olfactory systems into temporal signatures. However, the utilization of traditional pattern-matching methods and unpredictable shuffling of spikes in the rank-order impedes the performance of the system. In this paper, we present an …