教学课程资源库 更多>>

关键词： Computer engineering   Information technology   Web studies   Computer science  

摘要： A large enterprise with a critical reliance on Information Technology (IT) must manage a wide variety of connectivity products and protocols such as Local Area Networks (LAN) and Storage Area Networks (SAN). SANs and LANs intersect in large enterprises with multiple data centers but require different tools for similar tasks. Software Defined Networking (SDN) is a new and evolving set of techniques, tools, and standards intended to virtualize existing Internet Protocol (IP) networks and make their configurations more flexible and respond rapidly to a variety of requirements. SDN allows network engineers to simplify and automate network management with open protocols that can reduce cost. SANs are proprietary, expensive, and can be complex to configure and manage, which are many of the problems that engendered SDN for LANs, but SDN has not been extended to manage them. Some proprietary attempts to paint SANs as more SDN-like do not make them open, simple, and easier to integrate and automate, which are some key drivers of SDN. This paper investigates how SDN can be extended to manage SANs as well as LANs. SDN V1.2 added experimental match fields which enable switching SAN traffic in the same way as LAN traffic. Comparisons of current SAN management and methods using SDN experimenter match fields designed herein are shown. Using SDN to handle both SANs and LANs can address some issues of complexity and configuration to make it simpler to manage and automate these mixed environments as a whole. Using these modifications, SDN switches could manage both LANs and SANs, not as an overlay or a diluted proprietary solution, but as a simple, cheaper, open solution that is easier to deploy and automate. These modifications show that SDN enables innovation, one of its first driving factors.

关键词： Computer science  

摘要： Traumatic injuries are the leading cause of death in children under 18, with head trauma being the leading cause of death in children below 5. A large but unknown number of traumatic injuries are non-accidental, i.e. inflicted. The lack of sensitivity and specificity required to diagnose Abusive Head Trauma (AHT) from radiological studies results in putting the children at risk of re-injury and death. Modern Deep Learning techniques can be utilized to detect Abusive Head Trauma using Computer Tomography (CT) scans. Training models using these techniques are only a part of building AI-driven Computer-Aided Diagnostic systems. There are challenges in deploying the models to make them highly available and scalable. The thesis models the domain of Abusive Head Trauma using Deep Learning techniques and builds an AI-driven System at scale using best Software Engineering Practices. It has been done in collaboration with Phoenix Children Hospital (PCH). The thesis breaks down AHT into sub-domains of Medical Knowledge, Data Collection, Data Pre-processing, Image Generation, Image Classification, Building APIs, Containers and Kubernetes. Data Collection and Pre-processing were done at PCH with the help of trauma researchers and radiologists. Experiments are run using Deep Learning models such as DCGAN (for Image Generation), Pretrained 2D and custom 3D CNN classifiers for the classification tasks. The trained models are exposed as APIs using the Flask web framework, contained using Docker and deployed on a Kubernetes cluster. The results are analyzed based on the accuracy of the models, the feasibility of their implementation as APIs and load testing the Kubernetes cluster. They suggest the need for Data Annotation at the Slice level for CT scans and an increase in the Data Collection process. Load Testing reveals the auto-scalability feature of the cluster to serve a high number of requests.

关键词： information science \u0026 technology   computer science   information science \u0026 technology   computer science  

ISBN: (数字)9781681739915 ISBN: (纸本)9781681739908;9781681739922

摘要： With the explosive growth of mobile computing and Internet of Things (IoT) applications, as exemplified by AR/VR, smart city, and video/audio surveillance, billions of mobile and IoT devices are being connected to the Internet, generating zillions of bytes of data at the network edge. Driven by this trend, there is an urgent need to push the frontiers of artificial intelligence (AI) to the network edge to fully unleash the potential of IoT big data. Indeed, the marriage of edge computing and AI has resulted in innovative solutions, namely edge intelligence or edge AI. Nevertheless, research and practice on this emerging inter-disciplinary field is still in its infancy stage. To facilitate the dissemination of the recent advances in edge intelligence in both academia and industry, this book conducts a comprehensive and detailed survey of the recent research efforts and also showcases the authors" own research progress on edge intelligence. Specifically, the book first reviews the background and present motivation for AI running at the network edge. Next, it provides an overview of the overarching architectures, frameworks, and emerging key technologies for deep learning models toward training/inference at the network edge. To illustrate the research problems for edge intelligence, the book also showcases four of the authors" own research projects on edge intelligence, ranging from rigorous theoretical analysis to studies based on realistic implementation. Finally, it discusses the applications, marketplace, and future research opportunities of edge intelligence. This emerging interdisciplinary field offers many open problems and yet also tremendous opportunities, and this book only touches the tip of iceberg. Hopefully, this book will elicit escalating attention, stimulate fruitful discussions, and open new directions on edge intelligence.

关键词： Information technology   Computer science  

摘要： Microservice is a distinct architecture that exhibits a high degree of independence regarding the decentralization of development, operation, and teamwork. Decentralization is where different teams, processes, and procedures that are fragmented to promote quick, agile development. Many companies and development groups are moving away from building sizeable monolithic software projects and using the agility of microservices to decouple services and deliver fast solutions. Even with the advantages of fast, agile deployment and the decoupling extensive monolithic services, microservices architecture encompasses more than the decomposition of applications. If the organization and cultural impacts are not addressed initially and continually, the output of a microservices build may not match the desired goal. This dissertation analyzes the need for sharing or distribution of team structures, standardizations, and governance within a microservice development project. The study shows, with the addition of tools and or models utilized anywhere within a development System Development Lifecycle (SDLC) process, decentralized teams are more prepared to mitigate risks and distribute findings and governance between the teams. Last, the study shows, with the additions of models and tools, used by the microservice decentralized development teams, that they are more comfortable gathering data to help streamline their development process, making for a stronger cohesive build.

关键词： Mathematics   Artificial intelligence   Computer science  

摘要： With recent successes of recurrent neural networks (RNNs) for machine translation, and handwriting recognition tasks, we hypothesize that RNN approaches might be best suited for unsupervised anomaly detection in time series. In this thesis, we first contribute a comprehensive comparative evaluation of RNN-based deep learning methods for anomaly detection across a wide array of popular deep neural network architectures. In our second major contribution we observe that a key gap of deep learning based anomaly detection methods is the inability to identify portions of the data that led to the detected anomaly. To address this, we propose a novel explainability approach that aims to pinpoint regions of an input that lead to the detected anomaly. In sum, this thesis not only advances the state-of-the-art in deep learning based anomaly detection for time series data but it also contributes novel methods for producing explanations and evaluating explanation quality of anomaly detectors.

摘要： Retention of women through graduation in Computer Science (CS) majors is one of the biggest challenges for CS education. Most research in this area focuses on factors influencing attrition rather than why and how women remain committed. The goal of this research study is to understand retention from the perspective of women who persisted in their CS major. Using the theoretical lens of legitimate peripheral participation in communities of practice, I designed and conducted a study that involved focus groups, interviews, journey maps, and experience sampling methods. I found that retention of women in this study was influenced by four different types of interactions and eight different practices inside the CS major. I also found that learning was a matter of multimembership at the intersection of several different communities which supported both these women’s learning and retention. Finally, this dissertation provides a cross-case study narrative that highlights commonalities and differences of different pathways of ongoing participation investigated in this study. Such narrative is illustrated by five individual case studies of five women persisting in their CS major.

关键词： Computer science  

摘要： Human behavior recognition and analysis have been considered as a core technology that can facilitate a variety of applications. However, accurate detection and recognition of human behavior is still a big challenge that attracts a lot of research efforts. Among all the research works, motion sensors-based human behavior recognition is promising as it is low cost, low power, and easy to carry. In this dissertation, we use motion sensors to study human behaviors. First, we present Ultigesture (UG) wristband, a hardware platform for detecting and analyzing human behavior. The hardware platform integrates an accelerometer, gyroscope, and compass sensor, providing a combination of (1) fully open Application Programming Interface (API) for various application development, (2) appropriate form factor for comfortable daily wear, and (3) affordable cost for large scale adoption. Second, we study the hand gesture recognition problem when a user performs gestures continuously. we propose a novel continuous gesture recognition algorithm. It accurately and automatically separates hand movements into segments, and merges adjacent segments if needed, so that each gesture only exists in one segment. Then, we apply the Hidden Markov Model to classify each segment into one of predefined hand gestures. Experiments with human subjects show that the recognition accuracy is 99.4% when users perform gestures discretely, and 94.6% when users perform gestures continuously. Third, we study the hand gesture recognition problem when a user is moving. We propose a novel mobility-aware hand gesture segmentation algorithm to detect and segment hand gestures. We also propose a Convolutional Neural Network to classify hand gestures with mobility noises. For the leave-one-subject-out cross-validation test, experiments with human subjects show that the proposed segmentation algorithm achieves 94.0% precision, and 91.2% recall when the user is moving. The proposed hand gesture classification algorith

关键词： Electrical engineering   Computer science  

摘要： Many industries, including healthcare, are trying to take advantage of voice assistant systems by incorporating their technology into the industries' environment. However, not many companies or researchers have successfully integrated this technology into the daily practice of healthcare practitioners. Doctors, nurses and other healthcare practitioners spend much of their interaction time with patients clicking on the Electronic Medical Record (EMR) screen trying to access and update data. An important contribution of this research is to analyze this healthcare need for this technology in the healthcare practitioner’s workflow. This research developed an Alexa chatbot skill, "Doctor’s Assistant," as a generic application to help healthcare practitioners access and update EMR data via speech, while reducing data entry time and providing better patient care. The evaluation of this application illustrates that the "Doctor’s Assistant" skill is both effective and accurate.

关键词： Bones   Broadcasting   Hypercubes   Computer science   Roads   Fault tolerance   Independent spanning trees   alternating group networks   triangle breadth-first search   interconnection networks   Cayley graph  

摘要： A set of spanning trees in a graph G is called independent spanning trees (ISTs) if they are rooted at the same vertex r, and for each vertex v(not equal r) in G, the two paths from v to r in any two trees share no common vertex expect for v and r. ISTs can be applied in many research fields, such as fault-tolerant broadcasting and secure message distribution in reliable communication networks. Since Cayley graphs have been widely used to design interconnection networks, constructing ISTs on cayley graphs is worth studying. The alternating group network is a subclass of Cayley graphs, and the approach of constructing ISTs in alternating group networks is still unknown. In this paper, we propose a novel and simple top-down approach for constructing ISTs in alternating group networks. The main ideas of the algorithm are to use induction to develop small trees to big trees, to use a triangle breadth-first search (TBFS) process to create a backbone of an IST, and to use breadth-first search (BFS) process to connect the rest of nodes. Compared to other methods of different interconnection networks in the literature, the uniqueness of our method is that it does not need to determine the parent of one node by any rule, on the contrary others determine that by rules. The time complexity in n-dimensional alternating group network AN(n) is O(n(2) x n!), where nW is twice the number of nodes of ANn;hence it is polynomial time. We implement the algorithm in PHP and run cases from AN(3) to AN(10). The results reveal that all spanning trees of AN(3) to AN(10) are independent and that our algorithm is accurate and efficient.

关键词： Linguistics   Computer science  

摘要： The last decade was very fruitful in the field of subregular research. New classes of subregular languages and mappings were uncovered for modeling natural language phenomena, and new learning algorithms were developed for these classes. The subregular approach has been successfully applied to phonotactics (Heinz, 2010), rewrite processes in phonology and morphology (Chandlee, 2014), and even syntactic constraints over tree structures (Graf, 2018). However, the rapid pace of the theoretical research has not been matched when it comes to engineering considerations. Many of the proposed learning algorithms have not been implemented yet, and as a result, their performance on concrete data sets is not known. In my dissertation, I implement and experiment with some of the learners available for subregular languages and mappings. I test these learners on data that is modeled after linguistic phenomena such as word-final devoicing and various types of harmony systems. The code for these evaluations is available as part of my Python package SigmaPie (Aksenova, 2020). The findings of my thesis allow linguists and formal language theorists to assess possible applications of subregular techniques and approaches, in particular typology, cognitive science, and natural language processing.