教学课程资源库 更多>>

关键词： Computer science  

摘要： Substance misuse is a major problem in the world. In 2014, as many as 52,404 deaths in the US were caused by drug overdoses. In 2001, the monetary cost of drug misuse has been estimated to be 414 billion dollars. In this work, we explore the use of different machine learning algorithms in the prediction of cocaine misuse using structured and unstructured data found in electronic health records. These records contain various attributes that can help with this prediction, including but not limited to chart text data, previous diagnoses of certain diseases and information about the area the patient lives in. Traditional models using only one kind of data are compared with ensembles and neural networks. Finally, the models are evaluated using the PR AUC metric.

关键词： Computer science  

摘要： As a common medical imaging method, Computed Tomography (CT) can create tomographic images using X-ray data acquired from around the human body. However, high quality and adequately sampled X-ray measurement data are not always available. In this scenario, the tomographic image created by conventional reconstruction algorithms will be noisy, or contain artifacts. The goal of our study is to reconstruct high-quality tomographic images from noisy or incomplete scan data, including low-dose, sparse-view, and limited-angle scenarios, by utilizing novel deep learning techniques. In this project, we trained a Generative Adversarial Network (GAN) and used it as a signal prior in the Simultaneous Algebraic Reconstruction Technique (SART) iterative reconstruction algorithm. The GAN we trained includes a self-attention block to model long-range dependencies in the scan data. Compared with the state-of-the-art denoising cycle GAN, CIRCLE GAN, and a conventional mathematical reconstruction algorithm incorporating total variation minimization, our Self-Attention GAN for CT image reconstruction produces competitive results on solving limited-angle data reconstruction problems. On sparse view and low-dose scenarios, our method is not always the best among compared methods, but produces competitive results in some situations.

关键词： Science education   Gender studies   Higher education  

摘要： Institutions have implemented recruitment and retention initiatives in science, technology, engineering, and mathematics (STEM) degree programs; however, gender disparity of women in engineering and computer science programs persists. The purpose of this qualitative phenomenological study was to explore the lived experiences of female graduates from engineering and computer science programs. The conceptual framework was the theory of grit to explore how female students sustained their passion and perseverance through obstacles and adverse situations. Data were gathered through semistructured interviews with 17 female participants who graduated from engineering and computer science programs in the United States. Data were analyzed through a priori coding and thematic analysis. Six themes were identified: (a) resilience and perseverance through challenges, (b) finding passion to focus drive and determination, (c) build a support system, (d) confidence and belief in abilities, (e) advocate for self and other women, and (f) hard work is necessary for success. Findings may be used to develop equitable practices for all students, to reduce the presence of bias and stereotypes, and to promote targeted implementation of mentorship opportunities for female students in STEM programs.

关键词： Robotics   Computer science  

摘要： Robot motion planning is an important problem for real-world robot applications. Recently, the separation of workspaces between humans and robots has been gradually fading, and there is strong interest in developing solutions where collaborative robots (cobots) can interact or work safely with humans in a shared space or in close proximity. When working with humans in real-world environments, the robots need to plan safe motions under uncertainty stemming from many sources such as noise of visual sensors, ambiguity of verbal instruction, and variety of human motions. In this thesis, we propose novel optimization-based and learning-based robot motion planning algorithms to deal with the uncertainties in real-world environments. To handle the input noise of visual cameras and the uncertainty of shape and pose estimation of surrounding objects, we present efficient probabilistic collision detection algorithms for Gaussian and non-Gaussian error distributions. By efficiently computing upper bounds of collision probability between an object and a robot, we present novel trajectory planning algorithms that guarantee that the collision probability at any trajectory point is less than a user-specified threshold. To enable human-robot interaction using natural language instructions, we present a mapping function from grounded linguistic semantics to the coefficients of the motion planning optimization problem. The mapping function considers task descriptions and motion-related constraints. For collaborative robots working with a human in close proximity, we present human intention and motion prediction algorithms for efficient task ordering and safe motion planning. The robot observes the human poses in real-time and predicts the future human motion based on the history of human poses. We also present an occlusion-aware robot motion planning algorithm that accounts for occlusion in the visual sensor data and uses learning-based techniques for trajectory planning. We highligh

关键词： Computer engineering   Computer science  

摘要： Blockchain serves a wide variety of purposes in automating secure transactions. It provides an immutable distributed database that is shared and recorded with transactions verified by miners and stored by all parties. Public blockchain allows anyone to join in maintaining the distributed records of the system as a whole. Smart contracts offer immutable functions to facilitate secure processing and performance of services. This dissertation aims to increase the efficiency and security of blockchain transactions in three ways. Pre-validation of transactions for two-parties to eliminate problems including where the receiver does not agree to the transaction, where the receiver is specified incorrectly, or where the receiver does not yet exist. In this solution, blockchain is used only to ratify and witness an agreement already signed by both parties. The second enhancement is a hybrid of a blockchain and a set of trusted third parties. Each party not yet on the blockchain must choose a representative from the trusted set to establish their existence. Trusted third parties also vouch for physical entities such as houses' existence, adding them to the blockchain. Once parties and entities are registered on the smart contract, subsequent transactions do not require the trusted parties. Third, a smart contract is proposed to provide a decentralized voting protocol that can be used to define contracts between multiple parties with multiple rounds of voting on project management. The example analyzed is fundraising for a project from multiple investors and multiple rounds of payments for which investors vote, contingent on meeting deadlines.

关键词： Robotics   Artificial intelligence   Computer science  

摘要： Supervised learning paradigm for training Deep Convolutional Neural Networks (DCNN) rests on the availability of large amounts of manually annotated images, which are necessary for training deep models with millions of parameters. In this thesis, we present novel techniques for mitigating the required manual annotation, by generating large object instance datasets through compositing textured 3D models onto commonly encountered background scenes to synthesize training images. The generated training data augmented with real world annotations outperforms models trained only on real data. Non-textured 3D models are subsequently used for keypoint learning and matching, and 3D object pose estimation from RGB images. The proposed methods showcase promising results with regards to generalization on new and standard benchmark datasets. In the final part of the thesis, we investigate how these perception capabilities can be leveraged and encoded in a spatial map, in order to enable an agent to successfully navigate towards a target object.

关键词： Computer Science   Computer Engineering   topological data analysis   persistent homology   image classification   protein description   persistent cycles   gene expression analysis   applied machine learning   convolutional neural network   shallow learning   simplicial complex   homology   feature vector  

摘要： The growing need to understand and process data has driven innovation in many disparate areas of data science. The computational biology, graphics, and machine learning communities, among others, are striving to develop robust and efficient methods for such analysis. In this work, we demonstrate the utility of topological data analysis (TDA), a new and powerful tool to understand the shape and structure of data, to these diverse areas. First, we develop a new way to use persistent homology, a core tool in topological data analysis, to extract machine learning features for image classification. Our work focuses on improving modern image classification techniques by considering topological features. We show that incorporating this information to supervised learning models allows our models to improve classification, thus providing evidence that topological signatures can be leveraged for enhancing some of the pioneering applications in computer vision. Next, we propose a topology based, fast, scalable, and parameter-free technique to explore a related problem in protein analysis and classification. On an initial simplicial complex built using constituent protein atoms and bonds, simplicial collapse is used to construct a filtration which we use to compute persistent homology. This is ultimately our signature for the protein-molecules. Our method, besides being scalable, shows sizable time and memory improvements compared to similar topology-based approaches. We use the signature to train a protein domain classifier and compare state-of-the-art structure-based protein signatures to achieve a substantial improvement in accuracy. Besides considering the intervals of persistent homology like our first two applications, some applications need to find representative cycles for them. These cycles, especially the minimal ones, are useful geometric features functioning as augmentations for the intervals in a purely topological barcode. We address the problem of computing these

ISBN: (纸本)9787302559146

关键词： Mechanical engineering   Computer science   Causation   Earthquakes   Markets   Time-series analysis  

摘要： Ever since Granger first proposed the idea of quantitatively testing the causal relationship between data streams, the endeavor of accurately inferring the causality in data and using that information to predict the future has not stopped. Artificial Intelligence (AI), by utilizing the massive amounts of data, helps to solve complex problems, whether they include the diagnosis and detection of disease through medical imaging, email spam detection, or self-driving vehicles. Perhaps, this thesis will be trivial in ten years from now. AI has pushed humankind to reach the next technological level in technology. Nowadays, among most machine leaning inquiries, statistical relationships are determined using correlation measures. By feeding data into machine learning algorithms, computers update the algorithm’s parameters iteratively by extracting and mapping features to learning targets until the correlation increases to a significant level to cease the training process. However, with the increasing developments of powerful AI, there is really a shortage of exploring causality in data. It is almost self-evident that ”correlation is not causality”. Sometimes, the strong correlation established between variables through machine learning can be absurd and meaningless. Providing insight into causality information through data, which most of the machine learning methods fall short to do, is of paramount importance. The subsequent chapters detail the four endeavors of studying causality in financial markets, earthquakes, animal/human brain signals, the predictivity of data sets. In Chapter 2, we further developed the concept of causality networks into a higher-order causality network. We applied these to financial data and tested their validity and ability to capture the system’s causal relationship. In next Chapter 3, We examined another type of time series-earthquakes. Violent seismic activities decimate people's lives and destroy entire cities and areas. This begs us to under

关键词： Information science   Computer science  

摘要： Many software engineering companies create x86 programs that run on the Windows operating system, and most use various types of software protection to safeguard their intellectual property. Software protection in the form of obfuscation is the most common technical means for securing programs against man-at-the-end (MATE) attacks, whereby the original code is transformed to make it more resilient against adversarial analysis. This research evaluates the effectiveness of both custom and commercial-tool obfuscation transformations against attacks where Windows-based symbolic execution is used. Our experimental approach analyzes the effectiveness of symbolic analysis against unprotected x86 programs versus those that have been transformed by custom software-based hardware abstraction (SBHA) and also those that are transformed by commercial obfuscators such as Themida and VMProtect. Results indicate that SBHA and obfuscation by commercial tools provide high effectiveness against x86-based symbolic analysis tools such as Angr. In addition, the research shows the relative weakness of recovered SBHA circuit netlists against adversarial analyzers such as Z3 that use satisfiability theory (SAT) methods.