Computer Sciences Commons^™

An Improved Lower Bound For Sparse Reconstruction From Subsampled Walsh Matrices, Jaroslaw Blasiok, Patrick Lopatto, Kyle Luh, Jake Marcinek, Shravas Rao

Computer Science Faculty Publications and Presentations

We give a short argument that yields a new lower bound on the number of uniformly and independently subsampled rows from a bounded, orthonormal matrix necessary to form a matrix with the restricted isometry property. We show that a matrix formed by uniformly and independently subsampling rows of an N ×N Walsh matrix contains a K-sparse vector in the kernel, unless the number of subsampled rows is Ω(KlogKlog(N/K)) — our lower bound applies whenever min(K,N/K) > logC N. Containing a sparse vector in the kernel precludes not only the restricted isometry property, but more generally the application of those matrices for …

Exploring Transformers As Compact, Data-Efficient Language Models, Clayton Fields, Casey Kennington

Computer Science Faculty Publications and Presentations

Large scale transformer models, trained with massive datasets have become the standard in natural language processing. The huge size of most transformers make research with these models impossible for those with limited computational resources. Additionally, the enormous pretraining data requirements of transformers exclude pretraining them with many smaller datasets that might provide enlightening results. In this study, we show that transformers can be significantly reduced in size, with as few as 5.7 million parameters, and still retain most of their downstream capability. Further we show that transformer models can retain comparable results when trained on human-scale datasets, as few as …

Tiny Language Models Enriched With Multimodal Knowledge From Multiplex Networks, Clayton Fields, Osama Natouf, Andrew Mcmains, Catherine Henry, Casey Kennington

Computer Science Faculty Publications and Presentations

Large transformer language models trained exclusively on massive quantities of text are now the standard in NLP. In addition to the impractical amounts of data used to train them, they require enormous computational resources for training. Furthermore, they lack the rich array of sensory information available to humans, who can learn language with much less exposure to language. In this study, performed for submission in the BabyLM challenge, we show that we can improve a small transformer model’s data efficiency by enriching its embeddings by swapping the learned word embeddings from a tiny transformer model with vectors extracted from a …

Convolution Neural Networks For Phishing Detection, Arun D. Kulkarni

Computer Science Faculty Publications and Presentations

Phishing is one of the significant threats in cyber security. Phishing is a form of social engineering that uses e-mails with malicious websites to solicitate personal information. Phishing e-mails are growing in alarming number. In this paper we propose a novel machine learning approach to classify phishing websites using Convolution Neural Networks (CNNs) that use URL based features. CNNs consist of a stack of convolution, pooling layers, and a fully connected layer. CNNs accept images as input and perform feature extraction and classification. Many CNN models are available today. To avoid vanishing gradient problem, recent CNNs use entropy loss function …

Multispectral Image Analysis Using Convolution Neural Networks, Arun D. Kulkarni

Computer Science Faculty Publications and Presentations

Machine learning (ML) techniques are used often to classify pixels in multispectral images. Recently, there is growing interest in using Convolution Neural Networks (CNNs) for classifying multispectral images. CNNs are preferred because of high performance, advances in hardware such as graphical processing units (GPUs), and availability of several CNN architectures. In CNN, units in the first hidden layer view only a small image window and learn low level features. Deeper layers learn more expressive features by combining low level features. In this paper, we propose a novel approach to classify pixels in a multispectral image using deep convolution neural networks …

Quantum Key-Length Extension, Joseph Jaeger, Fang Song, Stefano Tessaro

Computer Science Faculty Publications and Presentations

Should quantum computers become available, they will reduce the effective key length of basic secret-key primitives, such as blockciphers. To address this we will either need to use blockciphers with inherently longer keys or develop key-length extension techniques to amplify the security of a blockcipher to use longer keys.

We consider the latter approach and revisit the FX and double encryption constructions. Classically, FX was proven to be a secure key-length extension technique, while double encryption fails to be more secure than single encryption due to a meet-in-the-middle attack. In this work we provide positive results, with concrete and tight …

Testing Research Software: A Survey, Nasir U. Eisty, Jeffrey C. Carver

Computer Science Faculty Publications and Presentations

Background Research software plays an important role in solving real-life problems, empowering scientific innovations, and handling emergency situations. Therefore, the correctness and trustworthiness of research software are of absolute importance. Software testing is an important activity for identifying problematic code and helping to produce high-quality software. However, testing of research software is difficult due to the complexity of the underlying science, relatively unknown results from scientific algorithms, and the culture of the research software community.

Aims The goal of this paper is to better understand current testing practices, identify challenges, and provide recommendations on how to improve the testing process …

From Machine Learning To Deep Learning: A Comprehensive Study Of Alcohol And Drug Use Disorder, Banafsheh Rekabdar, David L. Albright, Haelim Jeong, Sameerah Talafha

Computer Science Faculty Publications and Presentations

This study aims to train and validate machine learning and deep learning models to identify patients with risky alcohol and drug misuse in a Screening, Brief Intervention, and Referral to Treatment (SBIRT) program. An observational cohort of 6978 adults was admitted in the western region of Alabama at three medical facilities between January and December of 2019. Data were cleaned and pre-processed using data imputation techniques and an augmented sampling data method. The primary analysis involved the multi-class classification of alcohol and drug misuse. Our study shows that accurate identification of alcohol and drug use screening instrument scores was best …

Hierarchical Structure Of Yso Clusters In The W40 And Serpens South Region: Group Extraction And Comparison With Fractal Clusters, Jia Sun, Robert A. Gutermuth, Hongchi Wang, Shuinai Zhang, Min Long

Computer Science Faculty Publications and Presentations

Young stellar clusters are believed to inherit the spatial distribution like hierarchical structures of their natal molecular cloud during their formation. However, the change of the structures between the cloud and the young clusters is not well constrained observationally. We select the W40–Serpens South region (∼7 × 9 pc²) of the Aquila Rift as a testbed and investigate hierarchical properties of spatial distribution of young stellar objects (YSOs) in this region. We develop a minimum spanning tree (MST) based method to group stars into several levels by successively cutting down edges longer than an algorithmically determined critical value. …

Deep Near-Infrared Survey Towards The W40 And Serpens South Region In The Aquila Rift: A Comprehensive Catalogue Of Young Stellar Objects, Min Long

Computer Science Faculty Publications and Presentations

Active star-forming regions are excellent laboratories for studying the origins and evolution of young stellar object (YSO) clustering. The W40–Serpens South region is such a region, and we compile a large near- and mid-infrared catalogue of point sources in it, based on deep near-infrared observations of Canada-France-Hawaii Telescope (CFHT) in combination with Two Micron All Sky Survey (2MASS), UKIRT Infrared Deep Sky Survey (UKIDSS), and Spitzer catalogues. From this catalogue, we identify 832 YSOs, and classify 15, 135, 647, and 35 of them to be deeply embedded sources, Class I YSOs, Class II YSOs, and transition disc sources, respectively. In …

A Simpler Machine Learning Model For Acute Kidney Injury Risk Stratification In Hospitalized Patients, Yirui Hu, Kunpeng Liu, Kevin Ho, David Riviello, Jason Brown, Alex R. Chang, Gurmukteshwar Singh, H. Lester Kirchner

Computer Science Faculty Publications and Presentations

Background: Hospitalization-associated acute kidney injury (AKI), affecting one-in-five inpatients, is associated with increased mortality and major adverse cardiac/kidney endpoints. Early AKI risk stratification may enable closer monitoring and prevention. Given the complexity and resource utilization of existing machine learning models, we aimed to develop a simpler prediction model. Methods: Models were trained and validated to predict risk of AKI using electronic health record (EHR) data available at 24 h of inpatient admission. Input variables included demographics, laboratory values, medications, and comorbidities. Missing values were imputed using multiple imputation by chained equations. Results: 26,410 of 209,300 (12.6%) inpatients developed AKI during …

“Pictures Are Easier To Remember Than Spellings!”: Designing And Evaluating Kidspic: A Graphical Image-Based Authentication Mechanism, Dhanush Kumar Ratakonda, Hoda Mehrpouyan, Jerry Alan Fails

Computer Science Faculty Publications and Presentations

Children encounter difficulties when they login to computers or websites because they have challenges remembering passwords. To improve children’s authentication, we conducted a series of formative studies with children (n = 8, ages 6–11) to understand their authentication practices with respect to a traditional text-based password and a new graphical picture-based password called KidsPic. The results obtained from these initial investigations, a security analysis of these authentication mechanisms, and participatory design sessions with children (ages 6–11) inspired design enhancements to KidsPic. We subsequently conducted a study comparing KidsPic to a traditional text-based authentication mechanism (n = …

Pushing Boundaries Of Co-Design By Going Online: Lessons Learned And Reflections From Three Perspectives, Jerry Alan Fails, Dhanush Kumar Ratakonda, Nitzan Koren, Salma Elsayed-Ali, Elizabeth Bonsignore, Jason Yip

Computer Science Faculty Publications and Presentations

The global COVID-19 pandemic made significant changes to our day-to-day lives, which impacted how we conduct research and design — including co-design. In this article, we present case studies from three different co-design groups that pushed the boundaries of traditional co-design, and conducted multiple co-design sessions (more than 150 total) over the last year and a half. The case studies for each team include: the transition to online co-design; the pros and cons of logistics and design tools utilized during the co-design sessions; and the advances, challenges, and surprises. We compare and contrast themes that emerged from the case studies …

Motion-Adjustable Neural Implicit Video Representation, Long Mai, Feng Liu

Computer Science Faculty Publications and Presentations

Implicit neural representation (INR) has been successful in representing static images. Contemporary image-based INR, with the use of Fourier-based positional encoding, can be viewed as a mapping from sinusoidal patterns with different frequencies to image content. Inspired by that view, we hypothesize that it is possible to generate temporally varying content with a single image-based INR model by displacing its input sinusoidal patterns over time. By exploiting the relation between the phase information in sinusoidal functions and their displacements, we incorporate into the conventional image-based INR model a phase-varying positional encoding module, and couple it with a phase-shift generation module …

Quantum Algorithms For Attacking Hardness Assumptions In Classical And Post‐Quantum Cryptography, J.-F. Biasse, X. Bonnetain, E. Kirshanova, A. Schrottenloher, Fang Song

Computer Science Faculty Publications and Presentations

In this survey, the authors review the main quantum algorithms for solving the computational problems that serve as hardness assumptions for cryptosystem. To this end, the authors consider both the currently most widely used classically secure cryptosystems, and the most promising candidates for post-quantum secure cryptosystems. The authors provide details on the cost of the quantum algorithms presented in this survey. The authors furthermore discuss ongoing research directions that can impact quantum cryptanalysis in the future.

Fairness In Information Access Systems, Michael D. Ekstrand, Anubrata Das, Robin Burke, Fernando Diaz

Computer Science Faculty Publications and Presentations

Recommendation, information retrieval, and other information access systems pose unique challenges for investigating and applying the fairness and non-discrimination concepts that have been developed for studying other machine learning systems. While fair information access shares many commonalities with fair classification, there are important differences: the multistakeholder nature of information access applications, the rank-based problem setting, the centrality of personalization in many cases, and the role of user response all complicate the problem of identifying precisely what types and operationalizations of fairness may be relevant.

In this monograph, we present a taxonomy of the various dimensions of fair information access and …

Snerf: Stylized Neural Implicit Representations For 3d Scenes, Thu Nguyen-Phuoc, Feng Liu, Lei Xiao

Computer Science Faculty Publications and Presentations

This paper presents a stylized novel view synthesis method. Applying state-of-the-art stylization methods to novel views frame by frame often causes jittering artifacts due to the lack of cross-view consistency. Therefore, this paper investigates 3D scene stylization that provides a strong inductive bias for consistent novel view synthesis. Specifically, we adopt the emerging neural radiance fields (NeRF) as our choice of 3D scene representation for their capability to render high-quality novel views for a variety of scenes. However, as rendering a novel view from a NeRF requires a large number of samples, training a stylized NeRF requires a large amount …

The Multisided Complexity Of Fairness In Recommender Systems, Nasim Sonboli, Robin Burke, Michael Ekstrand, Rishabh Mehrotra

Computer Science Faculty Publications and Presentations

Recommender systems are poised at the interface between stakeholders: for example, job applicants and employers in the case of recommendations of employment listings, or artists and listeners in the case of music recommendation. In such multisided platforms, recommender systems play a key role in enabling discovery of products and information at large scales. However, as they have become more and more pervasive in society, the equitable distribution of their benefits and harms have been increasingly under scrutiny, as is the case with machine learning generally. While recommender systems can exhibit many of the biases encountered in other machine learning settings, …

Deep Convolution Neural Networks For Image Classification, Arun D. Kulkarni

Computer Science Faculty Publications and Presentations

Deep learning is a highly active area of research in machine learning community. Deep Convolutional Neural Networks (DCNNs) present a machine learning tool that enables the computer to learn from image samples and extract internal representations or properties underlying grouping or categories of the images. DCNNs have been used successfully for image classification, object recognition, image segmentation, and image retrieval tasks. DCNN models such as Alex Net, VGG Net, and Google Net have been used to classify large dataset having millions of images into thousand classes. In this paper, we present a brief review of DCNNs and results of our …

Ethical Implications For Children’S Use Of Search Tools In An Educational Setting, Monica Landoni, Theo Huibers, Emiliana Murgia, Maria Soledad Pera

Computer Science Faculty Publications and Presentations

In the classroom, search tools enable students to access online resources. While these tools have many benefits in theory, in practice there are also ethical issues to consider. In this article, we discuss a number of ethics-related problems teachers are faced with and they need to find solutions for. Based on our own research experience developing and deploying information discovery tools for the classroom (both in a traditional classroom setting and on the Internet due to the ongoing outbreak of COVID-19), we share insights about ethics and the role of the expert-in-the-loop, teachers, both as co-design partners and liaisons between …

Extending Tensor Virtual Machine To Support Deep-Learning Accelerators With Convolution Cores, Yanzhao Wang, Fei Xie

Computer Science Faculty Publications and Presentations

Deep-learning accelerators are increasingly popular. There are two prevalent accelerator architectures: one based on general matrix multiplication units and the other on convolution cores. However, Tensor Virtual Machine (TVM), a widely used deep-learning compiler stack, does not support the latter. This paper proposes a general framework for extending TVM to support deep-learning accelerators with convolution cores. We have applied it to two well-known accelerators: Nvidia's NVDLA and Bitmain's BM1880 successfully. Deep-learning workloads can now be readily deployed to these accelerators through TVM and executed efficiently. This framework can extend TVM to other accelerators with minimum effort.

Sl-Cyclegan: Blind Motion Deblurring In Cycles Using Sparse Learning, Ali Syed Saqlain, Li-Yun Wang, Zhiyong Liu

Computer Science Faculty Publications and Presentations

In this paper, we introduce an end-to-end generative adversarial network (GAN) based on sparse learning for single image motion deblurring, which we called SL-CycleGAN. For the first time in image motion deblurring, we propose a sparse ResNet-block as a combination of sparse convolution layers and a trainable spatial pooler k-winner based on HTM (Hierarchical Temporal Memory) to replace non-linearity such as ReLU in the ResNet-block of SL-CycleGAN generators. Furthermore, we take our inspiration from the domain-to-domain translation ability of the CycleGAN, and we show that image deblurring can be cycle-consistent while achieving the best qualitative results. Finally, we perform extensive …

Vigilrx: A Scalable And Interoperable Prescription Management System Using Blockchain, Alixandra Taylor, Austin Kugler, Praneeth Babu Marella, Gaby G. Dagher

Computer Science Faculty Publications and Presentations

Achieving interoperability between healthcare providers is a major challenge. Current systems for managing prescription records suffer from data siloing, unnecessary record duplication, and slow record transfers. In many systems, patients do not retain control over their prescription data. Instead, they must use an intermediary to access or transfer their records. Furthermore, record transfers suffer from differing standards between providers, outdated communication methods, and information blocking. Solving these problems necessitates the creation of an interoperable prescription management system. Realizing such a system requires considering security, efficiency, scalability, and other challenges. Recent regulatory actions attempt to address these challenges, but fundamental issues …

Machine Learning Methods For Generating High Dimensional Discrete Datasets, Giuseppe Manco, Ettore Ritacco, Antonino Rullo, Domenico Saccà, Edoardo Serra

Computer Science Faculty Publications and Presentations

The development of platforms and techniques for emerging Big Data and Machine Learning applications requires the availability of real-life datasets. A possible solution is to synthesize datasets that reflect patterns of real ones using a two-step approach: first, a real dataset X is analyzed to derive relevant patterns Z and, then, to use such patterns for reconstructing a new dataset X' that preserves the main characteristics of X. This survey explores two possible approaches: (1) Constraint-based generation and (2) probabilistic generative modeling. The former is devised using inverse mining (IFM) techniques, and consists of generating a dataset satisfying given …

Rate Maximization In A Uav Based Full-Duplex Multi-User Communication Network Using Multi-Objective Optimization, Syed Muhammad Hashir, Sabyasachi Gupta, Gavin Megson, Ehsan Aryafar, Joseph Camp

Computer Science Faculty Publications and Presentations

In this paper, we study an unmanned-aerial-vehicle (UAV) based full-duplex (FD) multi-user communication network, where a UAV is deployed as a multiple-input–multiple-output (MIMO) FD base station (BS) to serve multiple FD users on the ground. We propose a multi-objective optimization framework which considers two desirable objective functions, namely sum uplink (UL) rate maximization and sum downlink (DL) rate maximization while providing quality-of-service to all the users in the communication network. A novel resource allocation multi-objective-optimization-problem (MOOP) is designed which optimizes the downlink beamformer, the beamwidth angle, and the 3D position of the UAV, and also the UL power of the …

Automatic Transformation Of Natural To Unified Modeling Language: A Systematic Review, Sharif Ahmed, Arif Ahmed, Nasir U. Eisty

Computer Science Faculty Publications and Presentations

Context: Processing Software Requirement Specifications (SRS) manually takes a much longer time for requirement analysts in software engineering. Researchers have been working on making an automatic approach to ease this task. Most of the existing approaches require some intervention from an analyst or are challenging to use. Some automatic and semi-automatic approaches were developed based on heuristic rules or machine learning algorithms. However, there are various constraints to the existing approaches to UML generation, such as restrictions on ambiguity, length or structure, anaphora, incompleteness, atomicity of input text, requirements of domain ontology, etc. Objective: This study aims to better understand …

Software Engineering Approaches For Tinyml Based Iot Embedded Vision: A Systematic Literature Review, Shashank Bangalore Lakshman, Nasir U. Eisty

Computer Science Faculty Publications and Presentations

Internet of Things (IoT) has catapulted human ability to control our environments through ubiquitous sensing, communication, computation, and actuation. Over the past few years, IoT has joined forces with Machine Learning (ML) to embed deep intelligence at the far edge. TinyML (Tiny Machine Learning) has enabled the deployment of ML models for embedded vision on extremely lean edge hardware, bringing the power of IoT and ML together. However, TinyML powered embedded vision applications are still in a nascent stage, and they are just starting to scale to widespread real-world IoT deployment. To harness the true potential of IoT and ML, …

Spoken Language Interaction With Robots: Recommendations For Future Research, Casey Kennington

Computer Science Faculty Publications and Presentations

With robotics rapidly advancing, more effective human–robot interaction is increasingly needed to realize the full potential of robots for society. While spoken language must be part of the solution, our ability to provide spoken language interaction capabilities is still very limited. In this article, based on the report of an interdisciplinary workshop convened by the National Science Foundation, we identify key scientific and engineering advances needed to enable effective spoken language interaction with robotics. We make 25 recommendations, involving eight general themes: putting human needs first, better modeling the social and interactive aspects of language, improving robustness, creating new methods …

Developers Perception Of Peer Code Review In Research Software Development, Nasir U. Eisty, Jeffrey C. Carver

Computer Science Faculty Publications and Presentations

Context Research software is software developed by and/or used by researchers, across a wide variety of domains, to perform their research. Because of the complexity of research software, developers cannot conduct exhaustive testing. As a result, researchers have lower confidence in the correctness of the output of the software. Peer code review, a standard software engineering practice, has helped address this problem in other types of software.

Objective Peer code review is less prevalent in research software than it is in other types of software. In addition, the literature does not contain any studies about the use of peer code …

Drones, Virtual Reality, And Modeling: Communicating Catastrophic Dam Failure, H. R. Spero, I. Vazquez-Lopez, K. Miller, R. Joshaghani, S. Cutchin, J. Enterkine

Computer Science Faculty Publications and Presentations

Dam failures occur worldwide and can be economically and ecologically devastating. Communicating the scale of these risks to the general public and decision-makers is imperative. Two-dimensional (2D) dam failure hydraulic models inform owners and floodplain managers of flood regimes but have limitations when shared with non-specialists. This study addresses these limitations by constructing a 3D Virtual Reality (VR) environment to display the 1976 Teton Dam disaster case study using a pipeline composed of (1) 2D hydraulic model data (extrapolated into 3D), (2) a 3D reconstructed dam, and (3) a terrain model processed from UAS (Uncrewed Airborne System) imagery using Structure …