Open Access. Powered by Scholars. Published by Universities.®

Physical Sciences and Mathematics Commons™

Open Access. Powered by Scholars. Published by Universities.®

2019

Discipline
Institution
Keyword
Publication
Publication Type
File Type

Articles 1111 - 1140 of 15927

Full-Text Articles in Physical Sciences and Mathematics

Ssrgd: Simple Stochastic Recursive Gradient Descent For Escaping Saddle Points, Zhize Li Dec 2019

Ssrgd: Simple Stochastic Recursive Gradient Descent For Escaping Saddle Points, Zhize Li

Research Collection School Of Computing and Information Systems

We analyze stochastic gradient algorithms for optimizing nonconvex problems. In particular, our goal is to find local minima (second-order stationary points) instead of just finding first-order stationary points which may be some bad unstable saddle points. We show that a simple perturbed version of stochastic recursive gradient descent algorithm (called SSRGD) can find an $(\epsilon,\delta)$-second-order stationary point with $\widetilde{O}(\sqrt{n}/\epsilon^2 + \sqrt{n}/\delta^4 + n/\delta^3)$ stochastic gradient complexity for nonconvex finite-sum problems. As a by-product, SSRGD finds an $\epsilon$-first-order stationary point with $O(n+\sqrt{n}/\epsilon^2)$ stochastic gradients. These results are almost optimal since Fang et al. [2018] provided a lower bound $\Omega(\sqrt{n}/\epsilon^2)$ for finding …

Go to article

Twenty Years Of Open Source Software: From Skepticism To Mainstream, Gregorio Robles, Igor Steinmacher, Paul Adams, Christoph Treude Dec 2019

Twenty Years Of Open Source Software: From Skepticism To Mainstream, Gregorio Robles, Igor Steinmacher, Paul Adams, Christoph Treude

Research Collection School Of Computing and Information Systems

Open source software (OSS) has conquered the software world. You can see it nearly everywhere, from Internet infrastructure to mobile phones to the desktop. In addition to that, although many OSS practices were viewed with skepticism 20 years ago, several have become mainstream in software engineering today: from development tools such as Git to practices such as modern code reviews.

Go to article

Strongly Secure Authenticated Key Exchange From Supersingular Isogenies, Xiu Xu, Haiyang Xue, Kunpeng Wang, Ho Man Au, Song Tian Dec 2019

Strongly Secure Authenticated Key Exchange From Supersingular Isogenies, Xiu Xu, Haiyang Xue, Kunpeng Wang, Ho Man Au, Song Tian

Research Collection School Of Computing and Information Systems

This paper aims to address the open problem, namely, to find new techniques to design and prove security of supersingular isogeny-based authenticated key exchange (AKE) protocols against the widest possible adversarial attacks, raised by Galbraith in 2018. Concretely, we present two AKEs based on a double-key PKE in the supersingular isogeny setting secure in the sense of CK+, one of the strongest security models for AKE. Our contributions are summarised as follows. Firstly, we propose a strong OW-CPA secure PKE, 2PKEsidh, based on SI-DDH assumption. By applying modified Fujisaki-Okamoto transformation, we obtain a [OW-CCA, OW-CPA] secure KEM, 2KEMsidh. Secondly, we …

Go to article

Efficient Meta Learning Via Minibatch Proximal Update, Pan Zhou, Xiao-Tong Yuan, Huan Xu, Shuicheng Yan, Jiashi Feng Dec 2019

Efficient Meta Learning Via Minibatch Proximal Update, Pan Zhou, Xiao-Tong Yuan, Huan Xu, Shuicheng Yan, Jiashi Feng

Research Collection School Of Computing and Information Systems

We address the problem of meta-learning which learns a prior over hypothesis from a sample of meta-training tasks for fast adaptation on meta-testing tasks. A particularly simple yet successful paradigm for this research is model-agnostic meta-learning (MAML). Implementation and analysis of MAML, however, can be tricky; first-order approximation is usually adopted to avoid directly computing Hessian matrix but as a result the convergence and generalization guarantees remain largely mysterious for MAML. To remedy this deficiency, in this paper we propose a minibatch proximal update based meta-learning approach for learning to efficient hypothesis transfer. The principle is to learn a prior …

Go to article

Evaluation Of Relationship Between Lead-Dust Loading, Lead-Dust Concentration, And Total Dust Loading Metrics Across Multiple Data Sets, Charles Bevington Dec 2019

Evaluation Of Relationship Between Lead-Dust Loading, Lead-Dust Concentration, And Total Dust Loading Metrics Across Multiple Data Sets, Charles Bevington

Capstone Experience

Lead-dust monitoring studies report values as either lead-dust loadings µg/ft2 or as lead-dust concentrations µg/g. It is rare for studies to report both metrics. When only lead-dust loading values are present, professionals require an approach to estimate lead-dust concentration values. A literature search identified five studies that contained raw data for both lead-dust loading and lead-dust concentration. An additional thirty-two studies had summary-statistics available for both lead-dust loading and lead-dust concentration. Studies with raw-data were used to develop an empirically-based loading to concentration statistical relationship. Raw data sets were critically evaluated to determine whether elimination or …

Go to article

Image Classification Using Fuzzy Fca, Niruktha Roy Gotoor Dec 2019

Image Classification Using Fuzzy Fca, Niruktha Roy Gotoor

Department of Computer Science and Engineering: Dissertations, Theses, and Student Research

Formal concept analysis (FCA) is a mathematical theory based on lattice and order theory used for data analysis and knowledge representation. It has been used in various domains such as data mining, machine learning, semantic web, Sciences, for the purpose of data analysis and Ontology over the last few decades. Various extensions of FCA are being researched to expand it's scope over more departments. In this thesis,we review the theory of Formal Concept Analysis (FCA) and its extension Fuzzy FCA. Many studies to use FCA in data mining and text learning have been pursued. We extend these studies to include …

Go to article

Astrodynamics Of The Next Generation Space Weather Prediction Mission, Mark Herring Dec 2019

Astrodynamics Of The Next Generation Space Weather Prediction Mission, Mark Herring

Doctoral Dissertations and Master's Theses

Accurate prediction of the solar wind properties, interplanetary magnetic field direction and various space weather phenomena becomes ever more important as our dependence on Earth orbiting spacecraft increases. Different solar wind drivers can lead both to enhancements and losses of relativistic electrons in the outer radiation belts, thus posing a major risk to satellites. To further our understanding of the Sun’s impact on the near Earth space environment, as well as to provide predictive capabilities, a mission placing monitoring satellites in key orbits in the inner Solar System is being proposed. As part of that effort, the possibility of using …

Go to article

Generating Electromagnetic Nonuniformly Correlated Beams, Milo W. Hyde Iv, Xifeng Xiao, David G. Voelz Dec 2019

Generating Electromagnetic Nonuniformly Correlated Beams, Milo W. Hyde Iv, Xifeng Xiao, David G. Voelz

Faculty Publications

We develop a method to generate electromagnetic nonuniformly correlated (ENUC) sources from vector Gaussian Schell-model (GSM) beams. Having spatially varying correlation properties, ENUC sources are more difficult to synthesize than their Schell-model counterparts (which can be generated by filtering circular complex Gaussian random numbers) and, in past work, have only been realized using Cholesky decomposition—a computationally intensive procedure. Here we transform electromagnetic GSM field instances directly into ENUC instances, thereby avoiding computing Cholesky factors resulting in significant savings in time and computing resources. We validate our method by generating (via simulation) an ENUC beam with desired parameters. We find the …

Go to article

Climate Change Impacts On Phosphorus Loads In The Upper And Middle Charles River Watershed With Hspf Modeling, Meagan Riley Dec 2019

Climate Change Impacts On Phosphorus Loads In The Upper And Middle Charles River Watershed With Hspf Modeling, Meagan Riley

Graduate Masters Theses

Water quality in the Upper and Middle Charles River Watershed has improved over the past several decades primarily due to improvements statewide in wastewater management. However, climate change threatens this progress, with future projections promising increased precipitation and temperatures for the New England region. This study investigated the impact of climate change projections on total phosphorus loads in the Upper and Middle Charles River Watershed using the HSPF model. Model input data were extended through 2018 to update present day conditions represented by the previously calibrated and validated HSPF model. The updated model was then used to simulate the following …

Go to article

Machine Learning Models On Prognostic Outcome Prediction For Cancer Images With Multiple Modalities, Gengbo Liu Dec 2019

Machine Learning Models On Prognostic Outcome Prediction For Cancer Images With Multiple Modalities, Gengbo Liu

Theses and Dissertations

Machine learning algorithms have been applied to predict different prognostic outcomes for many different diseases by directly using medical images. However, the higher resolution in various types of medical imaging modalities and new imaging feature extraction framework brings new challenges for predicting prognostic outcomes. Compared to traditional radiology practice, which is only based on visual interpretation and simple quantitative measurements, medical imaging features can dig deeper within medical images and potentially provide further objective support for clinical decisions. In this dissertation, we cover three projects with applying or designing machine learning models on predicting prognostic outcomes using various types of …

Go to article

Predation Of Eastern Cottontail Rabbit (Sylvilagus Floridanus) By Great Blue Heron (Ardea Herodias), Carlos E. Cintra-Buenrostro, J. E. Cifuentes-Lujan Dec 2019

Predation Of Eastern Cottontail Rabbit (Sylvilagus Floridanus) By Great Blue Heron (Ardea Herodias), Carlos E. Cintra-Buenrostro, J. E. Cifuentes-Lujan

School of Earth, Environmental, and Marine Sciences Faculty Publications and Presentations

An opportunistic predation by Great Blue Heron (Ardea herodias) on Eastern cottontail rabbit (Sylvilagus floridanus) was observed in south Texas on 31 Oct 2019. The Great Blue Heron had already captured the Eastern cottontail rabbit at the first observation but the maneuvering process, killing, and ingestion were recorded photographically, which make this observation unique even though this might constitute the second report on Great Blue Heron eating Eastern cottontail rabbits.

Go to article

The Generation Of Operational Policy For Cyber-Physical Systems In Smart Homes, Jared Wayne Hall Dec 2019

The Generation Of Operational Policy For Cyber-Physical Systems In Smart Homes, Jared Wayne Hall

MSU Graduate Theses

The term “Cyber-Physical Systems” (CPS) refers to those systems which seamlessly integrate sensing, computation, control, and networking into physical objects and infrastructure [1]. In these systems, computers and networks of physical entities interact with each other to bring new capabilities to traditional physical systems. Since its introduction, the field of Cyber-Physical Systems (CPS) has evolved with new and interesting advancements concerning its capability, adaptability, scalability, and usability [1]. One such advancement is the unification of the Internet of Things (IoT), a concept that enables real-world everyday objects to connect to the internet and interact with each other, with CPS [1]. …

Go to article

Laser Spectroscopy Investigations On Molecular And Free Radicals Dynamics., Hamzeh Telfah Dec 2019

Laser Spectroscopy Investigations On Molecular And Free Radicals Dynamics., Hamzeh Telfah

Electronic Theses and Dissertations

Studying molecular dynamics and chemical kinetics is important to understand the chemical behavior of renewable energy sources. Laser spectroscopy techniques are powerful tools for the identification and diagnosis of such processes. In this dissertation, using our laser spectroscopy techniques, biofuels and solar energy were targeted as renewable energy sources study. We have studied the ultrafast exciton dynamics for a series of methylammonium lead bromide (CH3NH3PbBr3) nanostructures, nanocrystals (NCs, 0D), nanowires (NWs, 1D), and nanoplatelets (NPs, 2D) as a promising solar energy materials.1-2 Aided by analysis of UV−visible absorption and photoluminescence spectra, features in the transient absorption (TA) spectra are assigned …

Go to article

Using Property-Based Testing, Weighted Grammar-Based Generators, And A Consensus Oracle To Test Browser Rendering Engines And To Reproduce Minimized Versions Of Existing Test Cases, Joel David Martin Dec 2019

Using Property-Based Testing, Weighted Grammar-Based Generators, And A Consensus Oracle To Test Browser Rendering Engines And To Reproduce Minimized Versions Of Existing Test Cases, Joel David Martin

Computer Science and Engineering Theses

Verifying that a web browser rendering engine correctly renders all valid web pages is challenging due to the size of the input space (valid web pages), the difficulty of determining correct rendering for any given web page (the test oracle problem), and the degree to which normal variation in browser rendering behavior can obscure other differences (fonts, bor- ders, input controls, etc). These challenges lead to manual human involvement during the testing process. We propose a new Property-Based Testing (PBT) approach that addresses these challenges in order to enable automated web browser render testing. Our approach is composed of the …

Go to article

Towards Location Free Movement Recognition With Channel State Information, Chunhai Feng Dec 2019

Towards Location Free Movement Recognition With Channel State Information, Chunhai Feng

Computer Science and Engineering Dissertations

Channel state information based movement recognition has gathered immense attention over recent years. Different from traditional systems which usually require wearable sensors or surveillance cameras, many existing works achieved desirable performance with only wireless signals in various applications, including healthcare, security and Internet of Things, with different machine learning algorithms. However, it still remains many challenges to be solved. Particularly, the location dependent nature of channel state information is one of the most significant challenges remaining. Firstly, many previous researchers deploy and evaluate their systems with employing machine learning or deep neural networks. Because of the aforementioned challenge, the models …

Go to article

Detect Traffic Signs From Large Street View Images With Deep Learning, Zhifei Deng Dec 2019

Detect Traffic Signs From Large Street View Images With Deep Learning, Zhifei Deng

Computer Science and Engineering Theses

Autonomous driving is about to shaping the future of our life. Self-driving vehicles produced by Waymo or many other companies have demonstrated excellent driving capabilities on the road. However, accidents still happen. Correctly recognising the traffic signs, such as stop signs, is critical for a self-driving vehicle. Failing to recognise the traffic signs could lead to fatal accidents. Meanwhile, computer vision technology has made huge progress since the advent of deep learning, for example, image classification, object detection, and instance segmentation. Efforts have been made in developing faster and more accurate object detection methods. Faster R-CNN stands out as one …

Go to article

Parsing Code-Switched Taglish Language By Creating Constituents, Fadiah Qudah Dec 2019

Parsing Code-Switched Taglish Language By Creating Constituents, Fadiah Qudah

Computer Science and Engineering Theses

When extracting meaning from language, a common first step is to break down language into constituents, or words that work together as a unit. This task, known as parsing, typically follows a specific grammar in order decompose the language into its underlying structure composed of constituents. Difficulties with this grammar-based parsing occur, however, with real-world natural language due to its unstructured nature. Code-switching, the phenomenon of alternating between languages while communicating, further complicates this task by requiring us to parse based on two (or more) languages instead of one. In this thesis, a data-driven method to parse code-switched language into …

Go to article

Performance Modeling And Resource Provisioning For Data-Intensive Applications, Zhongwei Li Dec 2019

Performance Modeling And Resource Provisioning For Data-Intensive Applications, Zhongwei Li

Computer Science and Engineering Theses

Performance evaluation and resource provisioning are two most critical factors to be considered for designers of distributed systems at modern warehouse data centers. The ever-increasing volumes of data in recent years have pushed many businesses to move their computing tasks to the Cloud, which offers many benefits including the low system management and maintenance costs and better scalability. As a result, most recent prominently emerging workloads are data-intensive, calling for scaling out the workload to a large number of servers for parallel processing. Questions can be asked as what factors impact the system scaling performance, and how to efficiently schedule …

Go to article

Feature Extraction In Noise-Diverse Environments For Human Activities Recognition Using Wi-Fi, Sheheryar Arshad Dec 2019

Feature Extraction In Noise-Diverse Environments For Human Activities Recognition Using Wi-Fi, Sheheryar Arshad

Computer Science and Engineering Dissertations

With the rapid development of 802.11 standard and Internet of Things (IoT) applications, Wi-Fi (IEEE 802.11) has emerged as the most widely used wireless communication technology. Wi-Fi based sensing has found widespread use cases involving activity recognition, indoor localization, design of smart spaces and in healthcare applications. This dissertation presents the study of human activities’ sensing and recognition using channel state information (CSI) of Wi-Fi. We highlight the limitations of existing methods and consequently design the frameworks for collecting stable CSI and monitoring different indoor and outdoor environments for human activities. Specifically, this dissertation provide means to define and extract …

Go to article

Social Media Text Analysis Using Multi-Kernel Convolutional Neural Network, Anna Philips Dec 2019

Social Media Text Analysis Using Multi-Kernel Convolutional Neural Network, Anna Philips

Computer Science and Engineering Theses

Transportation planners and ride hailing platforms such as Uber and Lyft use their riders feedback to assess their services and monitor customer satisfaction. Social media websites such as Facebook, Instagram, LinkedIn and in particular Twitter provides a large dataset of micro-texts by users who regularly post to their social media accounts about their grievances with their ride experience. This data is often unorganized and intractable to process because of it’s extremely large size which is continuously increasing daily. In this project, we collected ride hailing service relevant text data from Twitter around New York and developed a novel Convolutional Neural …

Go to article

Approxml: Efficient Approximate Ad-Hoc Ml Models Through Materialization And Reuse, Faezeh Ghaderi Dec 2019

Approxml: Efficient Approximate Ad-Hoc Ml Models Through Materialization And Reuse, Faezeh Ghaderi

Computer Science and Engineering Theses

Machine Learning (ML) has become an essential tool in answering complex predictive analytic queries. Model building for large scale datasets is one of the most time-consuming parts of the data science pipeline. Often data scientists are willing to sacrifice some accuracy in order to speed up this process during the exploratory phase. In this report, we aim to demonstrate ApproxML, a system that efficiently constructs approximate ML models for new queries from previously constructed ML models using the concepts of model materialization and reuse. ApproxML supports a wide variety of ML models such as generalized linear models for supervised learning …

Go to article

Spatial Similarity Measures With Applications To Map Integration And Improving Accuracy Of Map Data Sets, Mousa Alhajlah Almotairi Dec 2019

Spatial Similarity Measures With Applications To Map Integration And Improving Accuracy Of Map Data Sets, Mousa Alhajlah Almotairi

Computer Science and Engineering Dissertations

These days we live in a digital era where most societies rely on applications that depend on digital data. One popular type of digital data that is the basis many of applications is spatial data. Road network maps are one of the spatial data sets that are available for many important applications. However, acquisition of Road Network maps is an expensive task in terms of cost and time, not to mention the maintenance and the updating costs on these spatial data sets. In addition, each Road network map is captured for specific applications such as: road navigation, topographic cartography for …

Go to article

Use Of Word Embedding To Generate Similar Words And Misspellings For Training Purpose In Chatbot Development, Sanjay Thapa Dec 2019

Use Of Word Embedding To Generate Similar Words And Misspellings For Training Purpose In Chatbot Development, Sanjay Thapa

Computer Science and Engineering Theses

The advancement in the field of Natural Language Processing and Machine Learning has played a significant role in the huge improvement of conversational Artificial Intelligence (AI). The use of text-based conversation AI such as chatbots have increased significantly for the everyday purpose to communicate with real people for a variety of tasks. Chatbots are deployed in almost all popular messaging platforms and channels. The rise of chatbot development frameworks based on machine learning is helping to deploy chatbot easily and promptly. These chatbot development frameworks use machine learning and natural language understanding (NLU) to understand users' messages and intents and …

Go to article

Comprehensive Study Of Generative Methods On Drug Discovery, Siyu Xiu Dec 2019

Comprehensive Study Of Generative Methods On Drug Discovery, Siyu Xiu

Computer Science and Engineering Theses

Observing the recent success of the deep learning (DL) technology in multiple life-changing application areas, e.g., autonomous driving, image/video search and discovery, natural language processing, etc., many new opportunities have presented themselves. One of the biggest ones lies in applying DL in accelerating the drug discovery, where millions of human lives could potentially be saved. However, applying DL into the drug discovery task turns out to be non-trivial. The most successful DL methods take fix-sized tensors/matrices, e.g., images, or sequences of tokens, e.g., sentences with variant numbers of words, as their inputs. However, none of these registers with the inputs …

Go to article

Distributed Deep Neural Networks Training For Brain Imaging Applications, Sudheer Raja Dec 2019

Distributed Deep Neural Networks Training For Brain Imaging Applications, Sudheer Raja

Computer Science and Engineering Theses

Over the recent years, Deep Neural Networks (DNNs) have surpassed human-level intelligence in recognizing and interpreting complex patterns in data. Ever since the ImageNet competition in 2012, Deep Learning (DL) has become a promising approach for solving numerous problems in the field of Computer Science. However, the neuroscience community is not able to utilize the DL algorithms effectively because the brain imaging datasets are huge in terms of size, and the current sequential training techniques do not scale up well for such big datasets. Without the proper amount of training data, training DNN models to competitive accuracies is quite challenging. …

Go to article

Deep Representation Learning For Clustering And Domain Adaptation, Mohsen Kheirandishfard Dec 2019

Deep Representation Learning For Clustering And Domain Adaptation, Mohsen Kheirandishfard

Computer Science and Engineering Dissertations

Representation learning is a fundamental task in the area of machine learning which can significantly influence the performance of the algorithms used in various applications. The main goal of this task is to capture the relationships between the input data and learn feature representations that contain the most useful information of the original data. Such representations can be further leveraged in many machine learning applications such as clustering, natural language analysis, recommender systems, etc. In this dissertation, we first present a theoretical framework for solving a broad class of non-convex optimization problems. The proposed method is applicable to various tasks …

Go to article

Deduplication-Aware Page Cache In Linux Kernel For Improved Read Performance, Venkata Satya Ravi Kiran Boggavarapu Dec 2019

Deduplication-Aware Page Cache In Linux Kernel For Improved Read Performance, Venkata Satya Ravi Kiran Boggavarapu

Computer Science and Engineering Theses

The amount of data being produced and consumed is increasing every day. As a result, there can be a large amount of redundant data in the storage system. Storing and accessing these duplicate data unnecessarily consumes disk space and I/O bandwidth. Deduplication techniques are widely deployed to remove the redundancy. In particular, the deduplication solutions that work at the block level are proven to be effective. These solutions aim to effectively use disk space and write bandwidth by avoiding duplicate data writes to the storage. However, such a design might not help in improving the read performance, which is critical …

Go to article

Hiding In Plain Sight? The Impact Of Face Recognition Services On Privacy, James Richard Ortega Dec 2019

Hiding In Plain Sight? The Impact Of Face Recognition Services On Privacy, James Richard Ortega

Computer Science and Engineering Theses

The public at large is increasingly concerned with privacy online. While the focus is on the data privately collected by platforms, there are also privacy concerns in the realm of public data. Seemingly innocuous information shared in public, on online platforms, can be pieced together to detrimentally affect one's privacy in unexpected ways. On YouTube there exists a rich public dataset for adversaries to analyze for the purposes of breaching privacy; particularly due to the intersection of location and facial data. The goal of this work is to characterize the privacy risks that exists on YouTube, and explore the viability …

Go to article

The Impact Of Toxic Replies On Twitter Conversations, Nazanin Salehabadi Dec 2019

The Impact Of Toxic Replies On Twitter Conversations, Nazanin Salehabadi

Computer Science and Engineering Theses

Social media has become an empowering agent for individual voices and freedom of expression. Yet, it can also serve as a breeding ground for hate speech. According to a Pew Research Center study, 41% of Americans have been personally subjected to harassing behavior online, 66% have witnessed these behaviors directed at others, and 18% have been subjected to particularly severe forms of harassment online, such as physical threats, harassment over a sustained period, sexual harassment, or stalking. Recently, many research studies have tried to understand online hate speech and its implications, focusing on detecting and characterizing hate speech. One limitation …

Go to article

Three-Dimensional Analytical Model Of Tidal Flow In The Damariscotta River Estuary, Me, Stephanie L. Ayres Dec 2019

Three-Dimensional Analytical Model Of Tidal Flow In The Damariscotta River Estuary, Me, Stephanie L. Ayres

Electronic Theses and Dissertations

Estuaries are coastal bodies of water subjected to strong tidal influence and characterized by their morphology, tidal dynamics, topography, and stratification. Tidal flow is critically important to the water circulation, nutrient influx, and sediment transport in or out of an estuary. However, tidal asymmetry enhanced by estuary shape and nonlinear processes can lead to complications in estuarine flow. Analytical models are used to systematically study tidal flow within an estuary. Previous studies have derived analytical models of varying complexity and applied them to investigate tidal and residual flow. This thesis derives a three-dimensional analytical model with a perturbation expansion of …

Go to article

First Prev Next Last