教学课程资源库 更多>>

关键词： Data collection   Wireless sensor networks   Systematics   Cloud computing   Bibliographies   Computer science   Information technology   Data collection   sensor cloud   the Internet of Things (IoT) wireless sensor networks (WSN)   systematic literature review (SLR)  

摘要： The integration of cloud computing and Wireless Sensor Networks (WSNs) to create Sensor-cloud helps in extending the data processing capability and storage capability of WSNs. Knowing how weak WSNs are with regards to communication ability, how to collect and upload sensory data to the cloud in limited time has become an issue in Sensor-cloud. In the last decade, with increasing interest by researchers in the domain, a considerable amount of research works have been conducted and published in the research domain. The main objective of this study is to systematically review the current research on data collection in Sensor-cloud. Hence, the study also aims at identifying, categorizing, and synthesizing important studies in the field of study. Accordingly, an evidence-based methodology is utilized in this study. By doing so, 43 relevant studies were identified and retrieved to answer the formulated research questions. The systematic methodology offers a methodical and rigorous study selection and evaluation process that is repeatable and precise. The result shows that research on data collection in Sensor-cloud is relatively consistent with stable output in the last five years. Ten proposal contributions were identified with System, Framework, and Algorithm being the most used by the selected studies. In conclusion, key research challenges and future research directions were identified and discussed for researchers to propose effective solutions to the existing challenges. Although research on data collection in Sensor-cloud is gaining some traction in recent years, the works in the domain are not sufficient and concrete proposals are needed to improve data collection.

关键词： Computer engineering   Computer science  

摘要： Program size and complexity have dramatically increased over time. To reduce their workload, developers began to utilize package managers. These packages managers allow third-party functionality, contained in units called packages, to be quickly imported into a project. Due to their utility, packages have become remarkably popular. The largest package repository, npm, has more than 1.2 million publicly available packages and serves more than 80 billion package downloads per month. In recent years, this popularity has attracted the attention of malicious users. Attackers have the ability to upload packages which contain malware. To increase the number of victims, attackers regularly leverage a tactic called typosquatting, which involves giving the malicious package a name that is very similar to the name of a popular package. Users who make a typo when trying to install the popular package fall victim to the attack and are instead served the malicious payload. The consequences of typosquatting attacks can be catastrophic. Historical typosquatting attacks have exported passwords, stolen cryptocurrency, and opened reverse shells. This thesis focuses on typosquatting attacks in package repositories. It explores the extent to which typosquatting exists in npm and PyPI (the de facto standard package repositories for *** and Python, respectively), proposes a practical defense against typosquatting attacks, and quantifies the efficacy of the proposed defense. The presented solution incurs an acceptable temporal overhead of 2.5% on the standard package installation process and is expected to affect approximately 0.5% of all weekly package downloads. Furthermore, it has been used to discover a particularly high-profile typosquatting perpetrator, which was then reported and has since been deprecated by npm. Typosquatting is an important yet preventable problem. This thesis recommends package creators to protect their own packages with a technique called defensive typosquat

关键词： Medical sciences   Computer science   Statistics   Medicine--Research   Inference   Clinical trials--Statistical methods  

摘要： Causal inference—the process of drawing a conclusion about the impact of an exposure on an outcome—is foundational to biomedicine, where it is used to guide intervention. The current gold-standard approach for causal inference is randomized experimentation, such as randomized controlled trials (RCTs). Yet, randomized experiments, including RCTs, often enforce strict eligibility criteria that impede the generalizability of causal knowledge to the real world. Observational data, such as the electronic health record (EHR), is often regarded as a more representative source from which to generate causal knowledge. However, observational data is non-randomized, and therefore causal estimates from this source are susceptible to bias from confounders. This weakness complicates two central tasks of causal inference: the replication or evaluation of existing causal knowledge and the generation of new causal knowledge. In this dissertation I (i) address the feasibility of observational data to replicate existing causal knowledge and (ii) present new methods for the generation of causal knowledge with observational data, with a focus on the causal tasks of comparing an outcome between two cohorts and the estimation of attributable risks of exposures in a causal system.

关键词： Computer science  

摘要： Data sources, e.g., Yelp or Twitter, that produce records without pre-defined schema (semi-structured) are popular nowadays. However, semi-structured data can be unwieldy for data exploration purposes due to an extremely large number of records as well as an accumulative number of attributes. Finding a more compact data representation (i.e., summary encoding) is thus critical before subsequent data analysis can be applied. Designing an appropriate summary encoding requires a trade-off between compactness and fidelity. In this thesis, a framework is developed for reasoning about the trade-off between summary compactness and fidelity. A measure of summary fidelity is proposed accordingly. Analytic and experimental evidences have been presented, which show that the proposed fidelity measure is not only efficiently computable, but also a meaningful measure of summary quality. To efficiently construct a high-fidelity summary encoding given a limitation on its verboseness, a clustering-based approach is identified and experiment results show that it produces results orders of magnitude faster and competitive with more powerful techniques for compression and summarization.

关键词： Computer engineering   Computer science  

摘要： An identity is a word or phrase that is used to refer to a particular social group, category, or role which holds a particular set of culturally-constructed meanings, or stereotypes. Most scholarship assumes that a well defined set of identities exist, which can then be used to study the implications of holding a particular identity. However, the identities we express for ourselves and others are often ill-defined. New methods and data are needed to acknowledge complex identities that manifest in the ever evolving domain of social media and to study the downstream impacts of using them to define oneself or another person. This thesis aims to extract and analyze the identities commonly used in social media. For this purpose, ten million biographies of Twitter users are used as the dataset and unigrams are extracted that pertain to identities and preferences. A similarity based network of identities is then generated to identify, study and explain clusters of related identities. The results indicate that a wide range of identities are used on social media, and that these identities cluster together in previously unexplored ways. Further, overlaps between these clusters signify the intricate nature and use of different expressions of identity in ambiguous contexts present in social media. Furthermore, the identity clusters indicate an expected trend of people choosing to identify with activities, attributes, occupations and preferences that they appear to define themselves by. This fragmentation of self identity is in accordance with prevailing literature in the field, but extends our understanding of how the self is presented in the context of social media.

关键词： Meteorology   Artificial intelligence   Computer science  

摘要： Downscaling is a landmark task in climate science and meteorology in which the goal is to use coarse scale, spatio-temporal data to infer values at finer scales. Statistical downscaling aims to approximate this task using statistical patterns gleaned from an existing dataset of downscaled values, often obtained from observations or physical models. In this work, we investigate the application of deep latent variable learning to the task of statistical downscaling. We present ClimAlign, a novel method for unsupervised, generative downscaling using adaptations of recent work in normalizing flows for variational inference. We evaluate the viability of our method using several different metrics on two datasets consisting of daily temperature and precipitation values gridded at low (1 degree latitude/longitude) and high (1/4 and 1/8 degree) resolutions. We show that our method achieves comparable predictive performance to existing supervised statistical downscaling methods while simultaneously allowing for both conditional and unconditional sampling from the joint distribution over high and low resolution spatial fields. We provide publicly accessible implementations of our method, as well as the baselines used for comparison, on GitHub.

关键词： Fluid mechanics   Information technology   Computer science  

摘要： This research study explores the effects of sending and downloading image and audio files through the WhatsApp platform. A better understanding of how images and audio are affected by WhatsApp is necessary because of its popularity and the prevalence of digital images and audio as evidence in digital forensic investigations. WhatsApp is a cross-platform communication service that allows the sending of media files and is one of the most popular services of its kind used worldwide. This application can be downloaded for use on mobile phones, iOS and Windows computers. There is also a website application available. The research done here will help investigators understand the process of different uploading and downloading techniques through different devices and the effects these methods have on file structure and metadata. An examination of recompression by WhatsApp, and how the WhatsApp software behaves when interacting with original images and audio will be conducted. The resulting media file’s structure, metadata, binary data, quantization table data and other compression characteristics will be examined, and changes between files that are sent and the corresponding files that are downloaded will be analyzed.

关键词： Artificial intelligence   Computer science  

摘要： Surveillance cameras are a great support in crime investigation and proximity alarms and play a vital role in public safety. However current surveillance systems require continuous human supervision for monitoring. The primary goal of the thesis is to prevent firearm-related violence and injuries. Automatic firearm detection enhances security and safety among people. Therefore, introducing a Deep Learning Object Detection model to detect Firearms and alert the corresponding police department is the main motivation. Visual Object Detection is a fundamental recognition problem in computer vision that aims to find objects of certain target classes with precise localization of input image and assign it to the corresponding label. However, there are some challenges arising to the wide variations in shape, size, appearance, and occlusions by the weapon carrier. There are other objections to the selection of best object detection model. So, three deep learning models are selected, explained and shown the differences in detecting the firearms. The dataset in this thesis is the customized selection of different categories of firearm collection like the pistol, revolver, handgun, bullet, rifle along with human detection. The entire dataset is annotated manually in pascalvoc format. Date augmentation technique has been used to enlarge our dataset and facilitate in detecting firearms that re deformed and having occlusion properties.. To detect firearms this thesis developed and practiced unified networks like SSD and two-stage object detectors like faster RCNN. SSD is easy to understand and detect objects however it fails to detect smaller objects. Faster RCNN are efficient and able to detect smaller firearms in the scene. Each class has attained more than 90% of confidence score.

关键词： Software   Correlation   Complexity theory   Software metrics   Education   Computer science   Process metrics   software defect prediction   software evolution  

摘要： Process metrics can reflect the software development process and the code changes which are the main causes of defects. So, recently, the researches have put more emphasis on process metrics in the field of software defect prediction. For evolving projects, it is more meaningful to study whether the software module introduces or eliminates defects or not, not whether the software module is defective or defect-free. However, no such work is available in the literature focusing on the change of defect state. Discovering the factors that influence the change of defect state in the process of software development can help us to understand the causes of software defects and improve the quality of subsequent software versions. Therefore, this paper presents an extensive empirical study on which process metrics are significantly important to change of defects in evolving projects. Five process metrics of 37 versions in 12 software projects are collected. We not only analyze the class correlation values and the classification performance values among five process metrics, but also perform statistical analysis to verify whether the experimental results are of practical value. The experimental results indicate that Number of Distinct Committers plays a significantly important role in the change of defect state, especially for elimination of defects, and Number of Revisions is the second, whereas Degree of Code Modification is the last. In addition, Average Number of Modified Lines is superior to Number of Modified Lines. Based on the experimental results, some suggestions for software development and software defect prediction are also discussed.

关键词： Computer science