Computer Sciences Commons^™

Fairer Evaluation Of Zero Shot Action Recognition In Videos, Kaiqiang Huang, Sarah Jane Delany, Susan Mckeever

Conference Papers

Zero-shot learning (ZSL) for human action recognition (HAR) aims to recognise video action classes that have never been seen during model training. This is achieved by building mappings between visual and semantic embeddings. These visual embeddings are typically provided via a pre-trained deep neural network (DNN). The premise of ZSL is that the training and testing classes should be disjoint. In the parallel domain of ZSL for image input, the widespread poor evaluation protocol of pre-training on ZSL test classes has been highlighted. This is akin to providing a sneak preview of the evaluation classes. In this work, we investigate …

Deep Learning Methods For Fingerprint-Based Indoor And Outdoor Positioning, Fahad Alhomayani

Electronic Theses and Dissertations

Outdoor positioning systems based on the Global Navigation Satellite System have several shortcomings that have deemed their use for indoor positioning impractical. Location fingerprinting, which utilizes machine learning, has emerged as a viable method and solution for indoor positioning due to its simple concept and accurate performance. In the past, shallow learning algorithms were traditionally used in location fingerprinting. Recently, the research community started utilizing deep learning methods for fingerprinting after witnessing the great success and superiority these methods have over traditional/shallow machine learning algorithms. The contribution of this dissertation is fourfold:

First, a Convolutional Neural Network (CNN)-based method for …

The Complexity Of Symmetry, Matthew Lemay

HMC Senior Theses

One of the main goals of theoretical computer science is to prove limits on how efficiently certain Boolean functions can be computed. The study of the algebraic complexity of polynomials provides an indirect approach to exploring these questions, which may prove fruitful since much is known about polynomials already from the field of algebra. This paper explores current research in establishing lower bounds on invariant rings and polynomial families. It explains the construction of an invariant ring for whom a succinct encoding would imply that NP is in P/poly. It then states a theorem about the circuit complexity partial …

Efficient Route Selection For Drone-Based Delivery Under Time-Varying Dynamics, Arindam Khanda, Federico Coro, Francesco Betti Sorbelli, Cristina M. Pinotti, Sajal K. Das

Computer Science Faculty Research & Creative Works

The use of drones can be a valuable solution for the problem of delivering goods for many reasons. In fact, they can be efficiently employed in time-critical situations when there is a traffic jam on the roads, to serve customers in hard-to-reach places, or simply to expand the business. However, due to limited battery capacities and the fact that drones can serve a single customer at a time, a drone-based delivery system (DBDS) aims to minimize the drones' energy usage for completing a route from the depot to the customer and go back to the depot for new deliveries. In …

Exploring Autoencoder-Based Error-Bounded Compression For Scientific Data, Jinyang Liu, Sheng Di, Kai Zhao, Sian Jin, Dingwen Tao, Xin Liang, Zizhong Chen, Franck Cappello

Computer Science Faculty Research & Creative Works

Error-bounded lossy compression is becoming an indispensable technique for the success of today's scientific projects with vast volumes of data produced during the simulations or instrument data acquisitions. Not only can it significantly reduce data size, but it also can control the compression errors based on user-specified error bounds. Autoencoder (AE) models have been widely used in image compression, but few AE-based compression approaches support error-bounding features, which are highly required by scientific applications. To address this issue, we explore using convolutional autoencoders to improve error-bounded lossy compression for scientific data, with the following three key contributions. (1) We provide …

Clustering Hypergraphs Via The Mapequation, Matthew Swan, Justin Zhan

Computer Science and Computer Engineering Faculty Publications and Presentations

A hypergraph is a generalization of a graph in that the restriction of pairwise affinity scores is lifted in favor of affinity scores that can be evaluated between an arbitrary number of inputs. Hypergraphs clustering is the process of finding groups in which members of a given hypergraph exhibit a high similarity and dissimilarity with members outside their group. In this paper, we generalize the well-known MapEquation, an optimization equation used in the clustering of nonhypergraphs, for hypergraphs. We develop an agglomerative algorithm, Hypergraph Random Walks (HRW), to find an approximate solution to the generalized MapEquation. Our algorithm requires neither …

Multidimensional Demographic Profiles For Fair Paper Recommendation, Reem Alsaffar, Susan Gauch

Computer Science and Computer Engineering Faculty Publications and Presentations

Despite double-blind peer review, bias affects which papers are selected for inclusion in conferences and journals. To address this, we present fair algorithms that explicitly incorporate author diversity in paper recommendation using multidimensional author profiles that include five demographic features, i.e., gender, ethnicity, career stage, university rank and geolocation. The Overall Diversity method ranks papers based on an overall diversity score whereas the Multifaceted Diversity method selects papers that fill the highest-priority demographic feature first. We evaluate these algorithms with Boolean and continuous-valued features by recommending papers for SIGCHI 2017 from a pool of SIGCHI 2017, DIS 2017 and IUI …

Wind Turbine Parameter Calibration Using Deep Learning Approaches, Rebecca Mccubbin

Electronic Theses and Dissertations

The inertia and damping coefficients are critical to understanding the workings of a wind turbine, especially when it is in a transient state. However, many manufacturers do not provide this information about their turbines, requiring people to estimate these values themselves. This research seeks to design a multilayer perceptron (MLP) that can accurately predict the inertia and damping coefficients using the power data from a turbine during a transient state. To do this, a model of a wind turbine was built in Matlab, and a simulation of a three-phase fault was used to collect realistic fault data to input into …

Detection Dns Tunneling Botnets, Bohdan Savenko, Sergii Lysenko, Kira Bobrovnikova, Oleg Savenko, George Markowsky

Computer Science Faculty Research & Creative Works

Botnets are often used in cyberattacks on network services and individual users, so the ability to detect botnets is very important. Botnets use DNS tunneling to send malicious command-and-control (CC) commands to victims' hosts. Unfortunately, DNS tunneling attacks are very hard to detect. The paper presents a new approach for DNS tunneling botnet detection, which considers all the features and architectural characteristics of botnets. The technique described in this paper is highly efficient at detecting DNS tunneling attacks.

Complex Interactions Between Multiple Goal Operations In Agent Goal Management, Sravya Kondrakunta

Browse all Theses and Dissertations

A significant issue in cognitive systems research is to make an agent formulate and manage its own goals. Some cognitive scientists have implemented several goal operations to support this issue, but no one has implemented more than a couple of goal operations within a single agent. One of the reasons for this limitation is the lack of knowledge about how various goals operations interact with one another. This thesis addresses this knowledge gap by implementing multiple-goal operations, including goal formulation, goal change, goal selection, and designing an algorithm to manage any positive or negative interaction between them. These are integrated …

Why Homogeneous Membranes Lead To Optimal Water Desalination: A Possible Explanation, Julio Urenda, Martine Ceberio, Olga Kosheleva, Vladik Kreinovich

Departmental Technical Reports (CS)

A recent experiment has shown that out of all possible biological membranes, homogeneous ones proved the most efficient water desalination. In this paper, we show that natural symmetry ideas lead to a theoretical explanation for this empirical fact.

Your 'Attention' Deserves Attention: A Self-Diversified Multi-Channel Attention For Facial Action Analysis, Xiaotian Li, Zhihua Li, Huiyuan Yang, Geran Zhao, Lijun Yin

Computer Science Faculty Research & Creative Works

Visual attention has been extensively studied for learning fine-grained features in both facial expression recognition (FER) and Action Unit (AU) detection. A broad range of previous research has explored how to use attention modules to localize detailed facial parts (e, g. facial action units), learn discriminative features, and learn inter-class correlation. However, few related works pay attention to the robustness of the attention module itself. Through experiments, we found neural attention maps initialized with different feature maps yield diverse representations when learning to attend the identical Region of Interest (ROI). In other words, similar to general feature learning, the representational …

Improving Lossy Compression For Sz By Exploring The Best-Fit Lossless Compression Techniques, Jinyang Liu, Sihuan Li, Sheng Di, Xin Liang, Kai Zhao, Dingwen Tao, Zizhong Chen, Franck Cappello

Computer Science Faculty Research & Creative Works

In the past decades, various lossy compressors have been studied broadly due to the ever-increasing volume of data being produced by today's scientific applications. SZ has been one of the best error-bounded lossy compressors ever raised, and it has a flexible framework that includes four adjustable steps: prediction, quantization, variable-length encoding, and lossless compression. In this paper, we improve the lossy compression performances of the SZ compression model by exploring different existing lossless compression techniques using the Squash data compression benchmark. Specifically, we first characterize the bytes outputted by the first three steps in SZ, then we investigate the best …

Statistical Inference: The Missing Piece Of Recsys Experiment Reliability Discourse, Ngozi Ihemelandu, Michael D. Ekstrand

Computer Science Faculty Publications and Presentations

This paper calls attention to the missing component of the recommender system evaluation process: Statistical Inference. There is active research in several components of the recommender system evaluation process: selecting baselines, standardizing benchmarks, and target item sampling. However, there has not yet been significant work on the role and use of statistical inference for analyzing recommender system evaluation results.

In this paper, we argue that the use of statistical inference is a key component of the evaluation process that has not been given sufficient attention. We support this argument with systematic review of recent RecSys papers to understand how statistical …

Incremental Unit Networks For Multimodal, Fine-Grained Information State Representation, Casey Kennington, David Schlangen

Computer Science Faculty Publications and Presentations

We offer a sketch of a fine-grained information state annotation scheme that follows directly from the Incremental Unit abstract model of dialogue processing when used within a multimodal, co-located, interactive setting. We explain the Incremental Unit model and give an example application using the Localized Narratives dataset, then offer avenues for future research.

A Practical And Secure Stateless Order Preserving Encryption For Outsourced Databases, Ning Shen, Jyh-Haw Yeh, Hung-Min Sun, Chien-Ming Chen

Computer Science Faculty Publications and Presentations

Order-preserving encryption (OPE) plays an important role in securing outsourced databases. OPE schemes can be either Stateless or Stateful. Stateful schemes can achieve the ideal security of order-preserving encryption, i.e., “reveal no information about the plaintexts besides order.” However, comparing to stateless schemes, stateful schemes require maintaining some state information locally besides encryption keys and the ciphertexts are mutable. On the other hand, stateless schemes only require remembering encryption keys and thus is more efficient. It is a common belief that stateless schemes cannot provide the same level of security as stateful ones because stateless schemes reveal the relative distance …

Adaptive Two-Stage Edge-Centric Architecture For Deeply-Learned Embedded Real-Time Target Classification In Aerospace Sense-And-Avoidance Applications, Nicholas A. Speranza

Browse all Theses and Dissertations

With the growing number of Unmanned Aircraft Systems, current network-centric architectures present limitations in meeting real-time and time-critical requirements. Current methods utilizing centralized off-platform processing have inherent energy inefficiencies, scalability challenges, performance concerns, and cyber vulnerabilities. In this dissertation, an adaptive, two-stage, energy-efficient, edge-centric architecture is proposed to address these limitations. A novel, edge-centric Sense-and-Avoidance architecture framework is presented, and a corresponding prototype is developed using commercial hardware to validate the proposed architecture. Instead of a network-centric approach, processing is distributed at the logical edge of the sensors, and organized as Detection and Classification Subsystems. Classical machine vision algorithms are …

Deep Learning For High-Impedance Fault Detection: Convolutional Autoencoders, Khushwant Rai, Firouz Badrkhani Ajaei, Farnam Hojatpanah, Katarina Grolinger

Electrical and Computer Engineering Publications

High-impedance faults (HIF) are difficult to detect because of their low current amplitude and highly diverse characteristics. In recent years, machine learning (ML) has been gaining popularity in HIF detection because ML techniques learn patterns from data and successfully detect HIFs. However, as these methods are based on supervised learning, they fail to reliably detect any scenario, fault or non-fault, not present in the training data. Consequently, this paper takes advantage of unsupervised learning and proposes a convolutional autoencoder framework for HIF detection (CAE-HIFD). Contrary to the conventional autoencoders that learn from normal behavior, the convolutional autoencoder (CAE) in CAE-HIFD …

Design Of Interactive Visualizations For Next-Generation Ultra-Large Communication Networks, Wenjun Chen, Anwar Haque, Kamran Sedig

Computer Science Publications

© 2013 IEEE. With the increasing size and complexity of next-generation communication networks, it is critical to utilize interactive visualizations to support the monitoring, planning, and management of networks. Effectively visualizing large-scale networks is difficult with traditional methods because of the high link density and complex node relationships. Given the limited screen space, to assist Internet Service Provider's (ISP) network planning and management activities, investigating how to present ultra-large-scale network data efficiently is crucial. This paper presents a real-Time interactive visualization system that combines the design strategies of progressive disclosure and multiple panels to elegantly visualize the large-scale networks and …

Going Meta On The Minimum Circuit Size Problem: How Hard Is It To Show How Hard Showing Hardness Is?, Zoë Bell

HMC Senior Theses

The Minimum Circuit Size Problem (MCSP) is a problem with a long history in computational complexity theory which has recently experienced a resurgence in attention. MCSP takes as input the description of a Boolean function f as a truth table as well as a size parameter s, and outputs whether there is a circuit that computes f of size ≤ s. It is of great interest whether MCSP is NP-complete, but there have been shown to be many technical obstacles to proving that it is. Most of these results come in the following form: If MCSP is NP-complete …

Exploring Cybersecurity Education At The K-12 Level, Weiru Chen, Yuming He, Xin Tian, Wu He, E. Langran (Ed.), D. Rutledge (Ed.)

Information Technology & Decision Sciences Faculty Publications

K-12 cybersecurity education is receiving growing attention with the growing number of cyberattacks and a shortage of cybersecurity professionals. However, there are many barriers for teachers to implement effective cybersecurity education in formal classroom environments. This study conducts a systematic literature review to examine the current state-of-the-art on K-12 cybersecurity education. Through the systematic literature review, we identified 20 closely relevant papers and recognized that a well-designed curriculum in cybersecurity education at the K-12 level is strongly needed to motivate students to pursue cybersecurity pathways and careers. The challenge and suggestions of curriculum design, teaching strategy, and learning assessment are …

Tagtools: Software Supporting Biologging Research, Sam Fynewever, Racheal Tejevbo, Stacy L. De Ruiter

Summer Research

Observing animals in obscure habitats has long posed a challenge in biology. Biologging addresses this broad problem, obtaining data about animals like their acceleration, GPS position, etc. using electronic tags. Data from tags can lead to important knowledge of animal behavior, as DeRuiter et al. (2013) have shown1 . However, data is often shared raw, causing inconsistencies. This project maintains TagTools, software addressing this specific problem. TagToolsfunctions (in Matlab, Octave & R) calibrate data and help analyze behavior. TagTools software is traditionally taught at in-person workshops. Our old documents (practicals) were hosted online, but had bugs & were very long, …

Designing Computer Applications To Model Geophysical Data For Siting Wells In Developing Countries, Aj Vrieland, Victor T. Norman

Summer Research

In some parts of the world, clean water is hard to find, and wells can be one of the best options available. However, creating a well is a big investment, so finding the best place to site the well is important. Commercial well siting software is prohibitively expensive for many development organizations, costing tens of thousands of dollars. Our open-source software duplicates some functionality of commercial software, but with none of the expense. This summer our project focused on software implementing two geophysical models, resistivity and seismic refraction.

Software As A Service: The Mediating Role Of Consequences Of Saas Diffusion On Firm Performance, Cristina Marie-Mccarthy Recchia

Wayne State University Dissertations

ABSTRACTSOFTWARE AS A SERVICE: THE MEDIATING ROLE OF CONSEQUENCES OF SAAS DIFFUSION ON FIRM PERFORMANCE by CRISTINA MARIE-MCCARTHY RECCHIA DECEMBER 2021 Advisor: Dr. Ratna Babu Chinnam Major: Industrial Engineering Degree: Doctor of Philosophy There are ample studies that support a positive link between information technology and firm performance. Bharadwaj (2000) and Chae (2014, 2018) are two examples that provided a foundation for this work. These scholars looked at how capabilities associated with information technology contribute to improved financial performance using a specific set of financial ratios. In addition, there are studies that examine a positive link between Software-as-a-Service (SaaS) and …

An Analysis Of C/C++ Datasets For Machine Learning-Assisted Software Vulnerability Detection, Daniel Grahn, Junjie Zhang

Computer Science and Engineering Faculty Publications

As machine learning-assisted vulnerability detection research matures, it is critical to understand the datasets being used by existing papers. In this paper, we explore 7 C/C++ datasets and evaluate their suitability for machine learning-assisted vulnerability detection. We also present a new dataset, named Wild C, containing over 10.3 million individual opensource C/C++ files – a sufficiently large sample to be reasonably considered representative of typical C/C++ code. To facilitate comparison, we tokenize all of the datasets and perform the analysis at this level. We make three primary contributions. First, while all the datasets differ from our Wild C dataset, some …

Recommending Collaborations Using Link Prediction, Nikhil Chennupati

Browse all Theses and Dissertations

Link prediction in the domain of scientific collaborative networks refers to exploring and determining whether a connection between two entities in an academic network may emerge in the future. This study aims to analyze the relevance of academic collaborations and identify the factors that drive co-author relationships in a heterogeneous bibliographic network. Using topological, semantic, and graph representation learning techniques, we measure the authors' similarities w.r.t their structural and publication data to identify the reasons that promote co-authorships. Experimental results show that the proposed approach successfully infer the co-author links by identifying authors with similar research interests. Such a system …

Promoting Diversity In Teaching Cybersecurity Through Gicl, Yuming He, Wu He, Xiaohong Yuan, Li Yang, Theo Bastiaens (Ed.)

Information Technology & Decision Sciences Faculty Publications

In summary, it is necessary to develop a diverse group of K-12 students’ interest and skills in cybersecurity as cyber threats continue to grow. Evidence shows that educating the next generation of cyber workers is a crucial job that should begin in elementary school. To ensure the effectiveness of cybersecurity education and equity at the K-12 level, teachers must create thoughtful plans for considering communities’ interests and needs, and to continually reconsider what’s working and how to adjust our strategies, approaches, design, and research plan to meet their specific needs, challenges, and strengths, particularly with students from under-served and underrepresented …

Free Water In T2 Flair Whitematter Hyperintensity Lesions, Xiaowei Yu, Norman Scheel, Lu Zhang, David C. Zhu, Rong Zhang, Dajiang Zhu

Computer Science Faculty Research & Creative Works

Background: White matter (WM) free water (FW) is likely associated with cerebral small vessel disease (CSVD). FWis the fraction of unconstrained water within an image voxel, which can be estimated from diffusion-weighted images. T2-weighted Fluid- Attenuated Inversion Recovery (FLAIR) whitematter hyperintensity (WMH)is awidely used index to assess the damages caused by CSVD. It is critical to characterize howFW content is altered inWMHlesions. In this work, we proposed a data processing framework to assess FW distributions in WMH and normal-appearingWM as well as in differentWMfiber tracts. Method: Single-shell diffusion-weighted image (SS-DWI) and T2 FLAIR image data of 133 cognitively normal (CN) …

Connectionless Edge-Cache Servers For Reducing Cellular Bandwidth Usage In Vehicular Networks, Rui Wang, Jayanthi Rao, Ce Zhou, Subir Biswas

Computer Science Faculty Research & Creative Works

This paper presents a novel caching mechanism based on Connectionless Edge Cache Servers in vehicular networks. The goal is to intelligently cache content within the vehicles and the edge servers so that majority of the vehiclerequested content can be obtained from those caches, thus minimizing the amount of cellular network usage needed for fetching content from a central server. A notable feature of the cache servers in this work is that they do not have backhaul connectivity. This makes the connectionless servers to be relatively less expensive compared to the usual Roadside Service Units (RSUs), and potentially moveable in response …

Lattice-Based Technique To Visualize And Compare Regional Terrorism Using The Global Terrorism Database, Linda Markowsky, George Markowsky

Computer Science Faculty Research & Creative Works

Order-theoretic lattices and their visualizations are proposed as a means of exploring and analyzing databases. The max-complete lattice is formed and the significance of lattice nodes and of the top node are demonstrated. These novel visualizations serve as a useful complement to well-known charting techniques such as bar charts and may extend the knowledge gained from critical databases. The Carver2 dataset is used as an illustrative example, highlighting the formation of the max-complete lattice from the original dataset and the computational advantage provided by compressing the lattice using only its irreducible elements. Regional information from the Global Terrorism Database (GTD), …