Search

Wireless capsule endoscopy is a modern non-invasive Internet of Medical Imaging Things that has been increasingly used in gastrointestinal tract examination. With about one gigabyte image data generated for a patient in each examination, automatic lesion detection is highly desirable to improve the efficiency of the diagnosis process and mitigate human errors. Despite many approaches for lesion detection have been proposed, they mainly focus on large lesions and are not directly applicable to tiny lesions due to the limitations of feature representation. As bleeding lesions are a common symptom in most serious gastrointestinal diseases, detecting tiny bleeding lesions is extremely important for early diagnosis of those diseases, which is highly relevant to the survival, treatment, and expenses of patients. In this article, a method is proposed to extract and fuse multi-scale deep features for detecting and locating both large and tiny lesions. A feature extracting network is first used as our backbone network to extract the basic features from wireless capsule endoscopy images, and then at each layer multiple regions could be identified as potential lesions. As a result, the features maps of those potential lesions are obtained at each level and fused in a top-down manner to the fully connected layer for producing final detection results. Our proposed method has been evaluated on a clinical dataset that contains 20,000 wireless capsule endoscopy images with clinical annotation. Experimental results demonstrate that our method can achieve 98.9% prediction accuracy and 93.5% score, which has a significant performance improvement of up to 31.69% and 22.12% in terms of recall rate and score, respectively, when compared to the state-of-the-art approaches for both large and tiny bleeding lesions. Moreover, our model also has the highest AP and the best medical diagnosis performance compared to state-of-the-art multi-scale models.

The growth rate of these enterprises has increased significantly in the last decade. Research has shown that over the past two decades, the amount of data has increased approximately tenfold every two years - this exceeded Moore's Law, which doubles the power of processors. About thirty thousand gigabytes of data are accumulated every second, and their processing requires an increase in the efficiency of data processing. Uploading videos, photos and letters from users on social networks leads to the accumulation of a large amount of data, including unstructured ones. This leads to the need for enterprises to work with big data of different formats, which must be prepared in a certain way for further work in order to obtain the results of modeling and calculations. In connection with the above, the research carried out in the article on processing and storing large data of an enterprise, developing a model and algorithms, as well as using new technologies is relevant. Undoubtedly, every year the information flows of enterprises will increase and in this regard, it is important to solve the issues of storing and processing large amounts of data. The relevance of the article is due to the growing digitalization, the increasing transition to professional activities online in many areas of modern society. The article provides a detailed analysis and research of these new technologies.

This research is about the nuisances of social media applications on a Wi-Fi network at a university campus in Ghana. The aim was to access the security risk on the network, the speed of the network, and the data consumption of those platforms on the network. Network Mapper (Nmap Zenmap) Graphical User Interface 7.80 application was used to scan the various social media platforms to identify the protocols, ports, services, etc. to enable in accessing the vulnerability of the network. Data consumption of users’ mobile devices was collected and analyzed. Device Accounting (DA) based on the various social media applications was used. The results of the analysis revealed that the network is prone to attacks due to the nature of the protocols, ports, and services on social media applications. The numerous users with average monthly data consumption per user of 4 gigabytes, 300 megabytes on social media alone are a clear indication of high traffic as well as the cost of maintaining the network. A URL filtering of the social media websites was proposed on Rockus Outdoor AP to help curb the nuisance.

Knowledge is an important aspect for every person; especially it is very important for physical education teachers, coaches, and athletes. Although the knowledge about nutrition, massage, and physiotherapeutic exercises was investigated in Burundi Country. This study was a descriptive lookup with blended methods. The participants of this lookup were 15 physical education teachers, coaches, and athletes taken by representative sampling. Data evaluation techniques used correlation and linear regression among indicators-variables, and between variables themselves with Software PLS-SEM and SPSS.21. The effects confirmed that there was a low correlation (r: 0.45) between the items-global knowledge, and the negative correlation (r = - 0.068) was found between prices of Gigabyte (GB) and source of learning confirmed that the more the price of gigabyte was expensive, the more the source of information about the learning was not sufficient. The correlation between nutrition and its indicators was not strong, consecutively presented (base: 0.339; components: o.355; knowledge: 0.402). The relation between learning (X1), nutrition (Y1) {Rx1y1: o.421}, and Knowledge (Y3) {Ry1y3} was not strong; the subjects have little knowledge about nutrition. The consecutively correlation of indicators (basics: 0.366; massage course: 0.378; knowledge: 0.441), on massage and physiotherapeutic exercises showed that subjects held little knowledge about the previous courses. Knowledge became strong management tools that help physical education, sport medicine, coaching training managers to decide how to improve peak of performance, to maintain the good quality of athletes and non-athletes: This learn about was the first to apply to evaluate career to the knowledge of physical education teachers, coaches, and athletes.

Current practices in scientific publishing are unsuitable for rapidly changing fields and for presenting updatable data sets and software tools. In this regard, and as part of our continuing pursuit of pushing scientific publishing to match the needs of modern research, we are delighted to announce the launch of GigaByte, an online open-access, open data journal that aims to be a new way to publish research following the software paradigm: CODE, RELEASE, FORK, UPDATE and REPEAT. Following on the success of GigaScience in promoting data sharing and reproducibility of research, its new sister, GigaByte, aims to take this even further. With a focus on short articles, using a questionnaire-style review process, and combining that with the custom built publishing infrastructure from River Valley Technologies, we now have a cutting edge, XML-first publishing platform designed specifically to make the entire publication process easier, quicker, more interactive, and better suited to the speed needed to communicate modern research.

Biological imaging tools continue to increase in speed, scale, and resolution, often resulting in the collection of gigabytes or even terabytes of data in a single experiment. In comparison, the ability of research laboratories to store and manage this data is lagging greatly. This leads to limits on the collection of valuable data and slows data analysis and research progress. Here we review common ways researchers store data and outline the drawbacks and benefits of each method. We also offer a blueprint and budget estimation for a currently deployed data server used to store large datasets from zebrafish brain activity experiments using light-sheet microscopy. Data storage strategy should be carefully considered and different options compared when designing imaging experiments.

Proper storage of data is a security necessity. This is because data can often contain valuable company information, which is liable to fall into the wrong hands if not properly managed. The problem is growing in tandem with increased content creation – the world is full of data. It is estimated that the aggregate amount of data, which doubles in size every two years, measures 4.4 zettabytes (trillion gigabytes) and is likely to reach a massive 44 zettabytes by 2020. 1 Companies are attempting to deal with a tidal wave of data. And a lack of integration inside a business can promote the dangerous phenomenon of content sprawl. This occurs when different departments do not harmonise their processes and there is no plan to address outdated content or to ensure that data is stored in the right way. Companies need to gain a tight rein on their digital assets and institute a rigid content management system to keep up with the data explosion. But that can be easier said than done, explains Paul Hampton of Alfresco.

The world is awash in data—by 2020 it is expected that there will be approximately 40 trillion gigabytes of data in existence, with that number doubling every 2 to 3 years. However, data production is not equal in all places—the global data landscape remains heavily concentrated on English-speaking, urban, and relatively affluent locations within the Global North. This inequality can contribute to new forms of digital and data colonialism. One partial solution to these issues may come in the form of crowdsourcing and volunteer geographic information (VGI), which allow Global South populations to produce their own data. Despite initial optimism about these approaches, many challenges and research gaps remain in understanding the opportunities and barriers that organizations endemic to the Global South face in carrying out their own sustainable crowdsourcing projects. What opportunities and barriers do these endemic organizations face when trying to carry out mapping projects driven by their own goals and desires? This paper contributes answers to this question by examining a VGI project that is currently mapping public libraries across the African continent. Our findings highlight how dramatically digital divides can bias crowdsourcing results; the importance of local cultural views in influencing participation in crowdsourcing; and the continued importance of traditional, authoritative organizations for crowdsourcing. These findings offer important lessons for researchers and organizations attempting to develop their own VGI projects in the Global South.

Big data management is a reality for an increasing number of organizations in many areas and represents a set of challenges involving big data modeling, storage and retrieval, analysis and visualization. However, technological resources, people and processes are crucial to facilitate the management of big data in any kind of organization, allowing information and knowledge from a large volume of data to support decision-making. Big data management can be supported by these three dimensions: technology, people and processes. Hence, this article discusses these dimensions: the technological dimension that is related to storage, analytics and visualization of big data; the human aspects of big data; and, in addition, the process management dimension that involves in a technological and business approach the aspects of big data management., volume 12, 2015

Edge intelligence (EI) has received a lot of interest because it can reduce latency, increase efficiency, and preserve privacy. More significantly, as the Internet of Things (IoT) has proliferated, billions of portable and embedded devices have been interconnected, producing zillions of gigabytes on edge networks. Thus, there is an immediate need to push AI (artificial intelligence) breakthroughs within edge networks to achieve the full promise of edge data analytics. EI solutions have supported digital technology workloads and applications from the infrastructure level to edge networks, however, there are still many challenges with the heterogeneity of computational capabilities and the spread of information sources. We propose a novel event-driven deep-learning framework, called EDL-EI (event-driven deep learning for edge intelligence), via the design of a novel event model by defining events using correlation analysis with multiple sensors in real-world settings and incorporating multi-sensor fusion techniques, a transformation method for sensor streams into images, and lightweight 2-dimensional convolutional neural network (CNN) models. To demonstrate the feasibility of the EDL-EI framework, we presented an IoT-based prototype system that we developed with multiple sensors and edge devices. To verify the proposed framework, we have a case study of air-quality scenarios based on the benchmark data provided by the USA Environmental Protection Agency for the most polluted cities in South Korea and China. We have obtained outstanding predictive accuracy (97.65% and 97.19%) from two deep-learning models on the cities’ air-quality patterns. Furthermore, the air-quality changes from 2019 to 2020 have been analyzed to check the effects of the COVID-19 pandemic lockdown.

Many researchers have recently used the prediction of protein secondary structure (local conformational states of amino acid residues) to test advances in predictive and machine learning technology such as Neural Net Deep Learning. Protein secondary structure prediction continues to be a helpful tool in research in biomedicine and the life sciences, but it is also extremely enticing for testing predictive methods such as neural nets that are intended for different or more general purposes. A complication is highlighted here for researchers testing their methods for other applications. Modern protein databases inevitably contain important clues to the answer, so-called "strong buried clues", though often obscurely; they are hard to avoid. This is because most proteins or parts of proteins in a modern protein data base are related to others by biological evolution. For researchers developing machine learning and predictive methods, this can overstate and so confuse understanding of the true quality of a predictive method. However, for researchers using the algorithms as tools, understanding strong buried clues is of great value, because they need to make maximum use of all information available. A simple method related to the GOR methods but with some features of neural nets in the sense of progressive learning of large numbers of weights, is used to explore this. It can acquire tens of millions and hence gigabytes of weights, but they are learned stably by exhaustive sampling. The significance of the findings is discussed in the light of promising recent results from AlphaFold using Google's DeepMind.

A machine learning model is private if it doesn't reveal (too much) about its training data. This three-part talk examines to what extent current networks are private. Standard models are not private. We develop an attack that extracts rare training examples (for example, individual people's names, phone numbers, or addresses) out of GPT-2, a language model trained on gigabytes of text from the Internet [2]. As a result there is a clear need for training models with privacy-preserving techniques. We show that InstaHide, a recent candidate, is not private. We develop a complete break of this scheme and can again recover verbatim inputs [1]. Fortunately, there exists provably-correct "differentiallyprivate" training that guarantees no adversary could ever succeed at the above attacks. We develop techniques to that allow us to empirically evaluate the privacy offered by such schemes, and find they may be more private than can be proven formally [3].

In this paper, a specific Android application for Bluetooth-based smart home system is presented. The aim of this research is to design, develop, and evaluate a user interface prototype for the smart home system. The designed mobile App is named by MINDS-apps V1, which is expected to be able to perform three tasks, (1) controlling by soft-control mode, i.e. an RGB ambient lamp and Fan; (2) controlling by hard-control, i.e. a generic power switch, curtain, door lock, and (3) monitoring purpose, i.e. humidity and temperature sensor. In total, there are six types of smart home devices used for the experiment. Using MIT App Inventor 2, the design process is divided into two phases: user interface design using the Components Designer and implementation of the programming logic using the Blocks Editor. Once the design is finished, the application is then compiled into debuggable APK file with 2.23 MB in size, after which it is tested on six aforementioned devices. The MINDS-apps is able to operate even in the low-end mobile device with 1 gigabyte of random access memory (RAM) and Bluetooth version 2.1.

Because of its low cost per gigabyte, hard disk drive (HDD)-based storage are still extensively used despite consuming more power than flash-based storage. In particular, HDDs can be effectively used as cold storage for energy efficiency by making some drives in spin down. However, typical cold storage cannot be used owing to high access latency, unless it is used for archival or backup purposes. Furthermore, it is difficult to apply many power-proportional solutions because they require reconfiguration of the server power domain and adjustments to the data layout. In this paper, we propose a serviceable cold storage design that can be applied gradually to online services. The proposed design only modifies the data server of a typical distributed storage system to let it utilize the spin-up or spin-down features of disk drives and determine the data location. Because the modified data server appears identical to the existing data nodes, it can be implemented in the same manner as the addition or removal of a data server. Our prototype is implemented on Ceph, a well-known distributed storage system, and its effectiveness in managing the skewed I/O pattern of applications is demonstrated using a benchmark that can reproduce the real I/O patterns of the LINE mobile messenger application.

The present information society creates huge quantities of textual information. This information explosion is being handled using Information Retrieval Systems. Their tasks are effective storage and searching in the text collections. The amount of text stored in IRS and auxiliary data structures constitute a suitable material for data compression. However, the data that form the textual database of every IRS are very mixed and it is therefore useful to study special data compression methods. This chapter focuses on high compression ratio algorithms specialized in text compression in IRS that enable a fast decompression of individual documents, fully integrated with the IRS, and work with an adequate compression speed. These methods uses word-based compression methods combined with topical development of input documents. Experimental results prove that clustering has a positive impact on the compression ratio. The advantage of implementing this approach is that it is not necessary to change the existing compression algorithm. The only thing that changes is the ordering in which compressed documents are input. Decompression algorithms are not influenced at all and knowledge of topical development is not necessary.

Anomaly detection in automotive systems has been a strong challenge: first, during the development phase, then after the manufacturing approval in ramp-up production and finally during the vehicles life cycle management. The numerous sensors positioned inside a vehicle generate more than a gigabyte of data at each second timeframe. These sensors are connected through the vehicle network, which comprises Electronic Control Units (ECUs) and Controller Area Network (CAN) buses. Each ECU gets input from its sensors, executes specific instructions and aims to monitor the vehicle's normal state detecting any irregular action corresponding to its observed behavior. The aggregator of all sensor data and control actions detects the anomalies in vehicle systems, which poses a multi-source big data problem. Detecting anomalies during manufacturing has turned out to be another research challenge after the introduction of Industry 4.0. This paper presents a performance comparison of different anomaly detection algorithms on time series originating from automotive sensor data. Interquartile range, isolation forest, particle swarm optimization and k-means clustering algorithms are used to detect outlier data in the study.

With over 117 million COVID-19–positive cases declared and the death count approaching 3 million, we would expect that the highly digitalized health systems of high-income countries would have collected, processed, and analyzed large quantities of clinical data from patients with COVID-19. Those data should have served to answer important clinical questions such as: what are the risk factors for becoming infected? What are good clinical variables to predict prognosis? What kinds of patients are more likely to survive mechanical ventilation? Are there clinical subphenotypes of the disease? All these, and many more, are crucial questions to improve our clinical strategies against the epidemic and save as many lives as possible. One might assume that in the era of big data and machine learning, there would be an army of scientists crunching petabytes of clinical data to answer these questions. However, nothing could be further from the truth. Our health systems have proven to be completely unprepared to generate, in a timely manner, a flow of clinical data that could feed these analyses. Despite gigabytes of data being generated every day, the vast quantity is locked in secure hospital data servers and is not being made available for analysis. Routinely collected clinical data are, by and large, regarded as a tool to inform decisions about individual patients, and not as a key resource to answer clinical questions through statistical analysis. The initiatives to extract COVID-19 clinical data are often promoted by private groups of individuals and not by health systems, and are uncoordinated and inefficient. The consequence is that we have more clinical data on COVID-19 than on any other epidemic in history, but we have failed to analyze this information quickly enough to make a difference. In this viewpoint, we expose this situation and suggest concrete ideas that health systems could implement to dynamically analyze their routine clinical data, becoming learning health systems and reversing the current situation.

The terms like Gigabyte and megabyte have become old and outdated now because with the improvement in the field of storage technology, the storage abilities have increased and cost of storage of data has also been reduced. In present times terms like petabyte, terabyte, etc are in vogue as the e-storage capacity has increased worldwide. But at the same time, with improvement in the storage technology, the world now has ever increasing volume of data, so the need for storage is on the rise.Nowadays, huge amount of data is being generated and updated within a short span of time. Therefore, to take care of its storage, access and further advancement, analysis is required. But since the vast amount of data is getting generated in short span of time so the analysis of data as well as management of data has become a tedious, hectic and a challenging job for the data centers and managers.Various issues and challenges are there while tackling data-related issues. For example, the data is available in different types such as structured, unstructured, semi-structured, heterogeneous, or homogeneous, which makes there sorting and storage difficult. Sometimes, the older tools are not sufficient to handle all these or a combination of one or more types in real-time, or in distributed systems such as cloud systems.The creation of data and the aggregation of data is approximately in zettabytes per year. Such, volume of data is not only the problem for data management, but also affects the attributes like variety, velocity, value, and complexity. Then, there are also some technical issues as well related to the data transportation and storage.Thus, we need to think about the issues of storage and management of ever-increasing e-data and develop the novel perspective and innovative ways to address these issues.In this paper, we have analyzed these issues and challenges by studying different methodologies for big data analysis and design for Sensing and omitting medical data, e-Health record, public health record and clinical data. This paper also includes the proposed framework related to distributed data ware house based computing that has been developed to handle a larger amount of data processing by supporting smart and cost effective medical support system.

Cloud providers offer a variety of storage solutions for hosting data, both in price and in performance. For Analytics and machine learning applications, object storage services are the go-to solution for hosting the datasets that exceed tens of gigabytes in size. However, such a choice results in performance degradation for these applications and requires extra engineering effort in the form of code changes to access the data on remote storage. In this paper, we present a generic end-to-end solution that offers seamless data access for remote object storage services, transparent data caching within the compute infrastructure, and data-aware topologies that boost the performance of applications deployed in Kubernetes. We introduce a custom-implemented cache mechanism that supports all the requirements of the former and we demonstrate that our holistic solution leads up to 48% improvement for Spark implementation of the TPC-DS benchmark and up to 191% improvement for the training of deep learning models from the MLPerf benchmark suite.

Industrial-scale discrete element simulations typically generate Gigabytes of data per time step, which implies that even opening a single file may require 5 - 15 minutes on conventional magnetic storage devices. Data science’s inherent multi-disciplinary nature makes the extraction of useful information challenging, often leading to undiscovered details or new insights. This study explores the potential of statistical learning to identify potential regions of interest for large scale discrete element simulations. We demonstrate that our in-house knowledge discovery and data mining system (KDS) can decompose large datasets into i) regions of potential interest to the analyst, ii) multiple decompositions that highlight different aspects of the data, iii) simplify interpretation of DEM generated data by focusing attention on the interpretation of automatically decomposed regions, and iv) streamline the analysis of raw DEM data by letting the analyst control the number of decomposition and the way the decompositions are performed. Multiple decompositions can be automated in parallel and compressed, enabling agile engagement with the analyst’s processed data. This study focuses on spatial and not temporal inferences.

A content delivery network (CDN) refers to a geographically distributed network of web servers, aiming to enhance the website performance and reliability, i.e., providing fast delivery of the Internet content. CDN is often part of a mature enterprise level network, but due to the distributed nature, CDN may be vulnerable to a distributed denial-of-service attack (DDoS), in which the cyber-attackers try to flood the victim with overloaded traffic or requests. In the literature, the impact of DDoS on CDN has not been widely studied. Motivated by this challenge, in this work, we emulate a CDN environment and investigate the effect of a particular DDoS attack on the environment. It is found that free CDN clusters could be destroyed by only several Gigabyte traffic. In the end, we also discuss some potential solutions to help defend against DDoS attacks.

It is evident that a huge amount of data is currently being generated. Across the world, 2.5 quintillion bytes of data is being recorded currently. It is almost equivalent to 0.5 Million TB or it is enough to occupy 10 Million Blue-ray disks. The amount of data is expected to surpass 44 trillion gigabytes at the end of 2020 (as compared to 4.4 trillion gigabytes during the end of 2013). The lion’s share of the data being recorded in the information systems is basically related to healthcare activities. Extracting useful information/insights from a large quantity of data is very important. Visualizing data could yield wonderful results, and summaries hidden in data, especially, visualization could do a wonderful job in health care. Data visualization saves time and conveys information more meaningfully. It is a powerful way to summarize which assists all stakeholders. This chapter presents an attempt to summarize healthcare data through exploratory data analysis and process mining control-flow discovery techniques. Exploratory data analysis of healthcare data presents a way to explore healthcare data meaningfully, and process mining based control flow visualization presents the way to extract the causal relationships between the activities of the process. Process mining way of visualizing healthcare helps in identifying the discrepancies between planned and actual healthcare processes. Final sections of this chapter present Process Mining based control flow visualizations on real-time event log detailed in healthcare information systems.

We propose the new multi-class of bagged proximal support vector machines (MC-Bag-PSVM) for handling the ImageNet challenging problem with very large number of images and a thousand classes. Our MC-Bag-PSVM trains in the parallel manner ensemble binary PSVM classifiers used for the One-Versus-All (OVA) multi-class strategy on multi-core computer with GPUs. The binary PSVM model is constructed by bagged binary PSVM models built in under-sampling training dataset. The numerical test results on ILSVRC 2010 dataset show that our MC-Bag-PSVM algorithm is faster and more accurate than the state-of-the-art linear SVM algorithm. An example of its effectiveness is given with an accuracy of 75.64% obtained in the classification of ImageNet-1000 dataset having 1,261,405 images in 2048 deep features into 1,000 classes in 29.5 min using a PC Intel(R) Core i7-4790 CPU, 3.6 GHz, 4 cores and Gigabyte GeForce RTX 2080Ti 11 GB GDDR6, 4352 CUDA cores.

Social media is a platform where people share their interests and opinions. Users posting their opinions about a product or a service they used is very common nowadays. This information can be important to marketers. Knowledge about how a product is perceived in a society is crucial to improve the sales and formulate marketing business strategies. The data flow in social media is of extreme high volume and this data is generated in a scale of hundreds of gigabytes each day. There are many applications which have been built to make the most use of this idle data one of the most important one being sentiment analysis.

Artificial intelligence is a general term that encompasses all means in which a computer uses modern methods to not just execute a specific action that a programmer has decided, but to teach itself how to conduct a task. Thus, the computer learns sometimes in the way a human might. In 2020, digital imaging sensors have matured and become advanced enough to capture images that far exceed that of film-based photography in the past. Not only has the quality of imaging increased, but also has the rate and ease at which imaging may be done such that it becomes feasible to acquire gigabytes of imagery within a single examination case. Deep learning is a class of machine learning that deals with the ability of a computer to learn, that is, to teach itself a new previously unknown task, as opposed to the methods which are designed and task-specific.

In modern communication systems, optical fiber transmission is widely used because of its low power consumption and wide frequency band. At the same time, by using the SFP (Small Form-factor Pluggable) module, the video transmission system will quickly build. Therefore, this paper uses the FPGA-integrated GTX (Gigabyte Transceiver) serial high-speed transceiver as the transmission and forwarding medium. The image is captured through the camera, then transmitted by the GTX transceiver and SPF fiber interface, and finally the video image is displayed by the HDMI interface. By the analysis of Vivado, the requirements of timing and layout that could be met.

Background Emerging neuroimaging datasets (collected with imaging techniques such as electron microscopy, optical microscopy, or X-ray microtomography) describe the location and properties of neurons and their connections at unprecedented scale, promising new ways of understanding the brain. These modern imaging techniques used to interrogate the brain can quickly accumulate gigabytes to petabytes of structural brain imaging data. Unfortunately, many neuroscience laboratories lack the computational resources to work with datasets of this size: computer vision tools are often not portable or scalable, and there is considerable difficulty in reproducing results or extending methods. Results We developed an ecosystem of neuroimaging data analysis pipelines that use open-source algorithms to create standardized modules and end-to-end optimized approaches. As exemplars we apply our tools to estimate synapse-level connectomes from electron microscopy data and cell distributions from X-ray microtomography data. To facilitate scientific discovery, we propose a generalized processing framework, which connects and extends existing open-source projects to provide large-scale data storage, reproducible algorithms, and workflow execution engines. Conclusions Our accessible methods and pipelines demonstrate that approaches across multiple neuroimaging experiments can be standardized and applied to diverse datasets. The techniques developed are demonstrated on neuroimaging datasets but may be applied to similar problems in other domains.

Gigabyte Passive Optical Network (GPON) technology can be called one of the most promising and progressive among all available ways to access the Internet today. This network has the ability to meet the needs of people in the speed of information exchange. It also meets modern requirements and has the potential capabilities for future development.

Data analysis and interpretation have now taken center stage with the advent of the digital age and the sheer amount of data can be frightening. In fact, a Digital Universe study found that the total data supply in 2012 was 2.8 trillion gigabytes! Based on that amount of data alone, it is clear the calling card of any successful enterprise in today's global world will be the ability to analyze complex data, produce actionable insights and adapt to new market needs all at the speed of thought. Business dashboards are the digital age tools for big data. Capable of displaying key performance indicators (KPIs) for both quantitative and qualitative data analyses, they are ideal for making the fast-paced and data-driven market decisions that push today's industry leaders to sustainable success. Through the art of streamlined visual communication, data dashboards permit businesses to engage in real-time and informed decision making, and are key instruments in data interpretation. This research article based on methods, benefits and problems of the interpretation of data.

Gaming is an important class of workloads for mobile devices. They are not only one of the biggest markets for game developers and app stores, but also amongst the most stressful applications for the SoC. In these workloads, much of the computation is user-driven, i.e. events captured from sensors drive the computation to be performed. Consequently, event processing constitutes the bulk of energy drain for these applications. To address this problem, we conduct a detailed characterization of event processing activities in several popular games and show that (i) some of the events are exactly repetitive in their inputs, not requiring any processing at all; or (ii) a significant number of events are redundant in that even if the inputs for these events are different, the output matches events already processed. Memoization is one of the obvious choices to optimize such behavior, however the problem is a lot more challenging in this context because the computation can span even functional/OS boundaries, and the input space required for tables can takes gigabytes of storage. Instead, our Selecting Necessary InPuts (SNIP) software solution uses machine learning to isolate the input features that we really need to track in order to considerably shrink memoization tables. We show that SNIP can save up to 32% of the energy in these games without requiring any hardware modifications.

As the core infrastructure of cloud data operation, exchange and storage, data centerneeds to ensure its security and reliability, which are the important prerequisites for the development of cloud computing. Due to various illegal accesses, attacks, viruses and other security threats, it is necessary to protect the boundary of cloud data center through security gateway. Since the traffic growing up to gigabyte level, the secure gateway must ensure high transmission efficiency and different network services to support the cloud services. In addition, data center is gradually evolving from IPv4 to IPv6 due to excessive consumption of IP addresses. Packet classification algorithm, which can divide packets into different specific streams, is very important for QoS, real-time data stream application and firewall. Therefore, it is necessary to design a high performance IPv6 packet classification algorithm suitable for security gateway.AsIPv6 has a128-bitIP address and a different packet structure compared with IPv4, the traditional IPv4 packet classification algorithm is not suitable properly for IPv6 situations. This paper proposes a fast packet classification algorithm for IPv6 - PCHA (packet classification based on hash andAdelson-Velsky-Landis Tree). It adopts the three flow classification fields of source IPaddress(SA), destination IPaddress(DA) and flow label(FL) in the IPv6 packet defined by RFC3697 to implement fast three-tuple matching of IPv6 packet. It is through hash matching of variable length IPv6 address and tree matching of shorter flow label. Analysis and testing show that the algorithm has a time complexity close to $O(1)$ in the acceptable range of space complexity, which meets the requirements of fast classification of IPv6 packetsand can adapt well to the changes in the size of rule sets, supporting fast preprocessing of rule sets. Our algorithm supports the storage of 500,000 3-tuple rules on the gateway device and can maintain 75% of the performance of throughput for small packets of 78 bytes.

This paper proposes a block solution method for the Extreme Learning Machine. It combines the speed of a direct non-iterative solver with minimal memory requirements. The method is suitable for edge computing scenarios running on a mobile device with GPU acceleration. The implementation tested on the GPU of iPad Pro outperforms a laptop CPU, and trains a 19,000-neuron model using under one gigabyte of memory. It confirms the feasibility of Big Data analysis on modern mobile devices.

One of the problems that has emerged during the intensive usage of the GOLDi remote lab over the years is the usage of external third-party development tools, necessary for the software- or hardware-oriented design. The installation and setup (e.g. of Atmel Studio or Intel Quartus Prime) on a private PC or laptop is quite complicated, requires some expert knowledge to use these tools safely and effectively, and is not always platform-independent. Above this, these IDE occupy several gigabytes in memory size and sometimes take hours to install. These very powerful tools, of which only a fraction of the offered functionality is needed to solve the given educational task, generate a large overhead. In addition, high school and university students may have only little or no knowledge of microcontrollers, FPGA, compiler linker toolchain, hardware-related languages, etc.

Persistent memory (PMEM) is an emerging byte-addressable memory device that sits on a memory bus like conventional DRAM. A first PMEM product, Intel® Optane™ DC Persistent Memory (DCPM), has a larger capacity and lower cost per gigabyte than DRAM, although its performance is lower than that of DRAM. Therefore, hybrid memory systems that combine DRAM and DCPM for main memory are recommended to take advantage of both of them. However, our previous work revealed significant unfairness between two processes co-running on a real hybrid memory system; the performance of a process accessing DRAM is significantly degraded by another performing frequent writes to DCPM, but not vice versa. In this work, we propose FairHym that is a dynamic frequency scaling technique to improve the inter-process fairness on hybrid memory systems. It decreases the operating frequencies of CPU cores that run a process performing frequent DCPM writes in order to throttle its access frequency and prevent DRAM accesses from being blocked. We implement FairHym as a user-level runtime system and evaluate it with 36 two-process workloads on a real server. The evaluation results show that FairHym improves a fairness metric from 0.66 to 0.86 on average, compared to the default setting that maximizes the frequencies of all cores.

Ubiquitous access and enhanced Internet speeds have paved ways for online educational reforms at a large scale. There has been a widespread adoption of modern educational applications, ranging from interactive applets, video lessons and online quizzes to remotely conducting laboratory experiments. Consequently, there is a demand to provision more bandwidth to satisfy the users expectations. In this paper, we propose an approach to enhance the Quality of Service (QoS) in a University campus network and efficiently utilize the available bandwidth. Typically within a University, some requests are similar e.g., operating system updates, Linux package installs, Python pip packages and others. These are huge data transfer requests ranging from Megabytes to Gigabytes, and consume a large amount of bandwidth on external access links to the Internet. Redundant requests of this nature from a large user base lead to enormous wastage of bandwidth. The proposed approach overcomes this concern by setting up a containerized forward proxy with a generic cache for popular traffic in the University. Our experiments on a live network at National Institute of Technology Karnataka, Surathkal show that a large number of redundant requests can be successfully served from this Virtualized Network Function (VNF), thereby enhancing the QoS and efficiently utilizing the available bandwidth. The proposed system is able to reduce the latency by over 60% and saves 34GB of data per day on an average. Although the proposed approach is tested in a University environment in this work, it is applicable for other caching requirements with minor modifications. Moreover, since this cache is implemented as a VNF, it is portable and easy to deploy.

Location-based social networks (LBSNs) have been studied extensively in recent years. However, utilizing real-world LBSN data sets yields several weaknesses: sparse and small data sets, privacy concerns, and a lack of authoritative ground-truth. To overcome these weaknesses, we leverage a large-scale LBSN simulation to create a framework to simulate human behavior and to create synthetic but realistic LBSN data based on human patterns of life. Such data not only captures the location of users over time but also their interactions via social networks. Patterns of life are simulated by giving agents (i.e., people) an array of “needs” that they aim to satisfy, e.g., agents go home when they are tired, to restaurants when they are hungry, to work to cover their financial needs, and to recreational sites to meet friends and satisfy their social needs. While existing real-world LBSN data sets are trivially small, the proposed framework provides a source for massive LBSN benchmark data that closely mimics the real-world. As such, it allows us to capture 100% of the (simulated) population without any data uncertainty, privacy-related concerns, or incompleteness. It allows researchers to see the (simulated) world through the lens of an omniscient entity having perfect data. Our framework is made available to the community. In addition, we provide a series of simulated benchmark LBSN data sets using different synthetic towns and real-world urban environments obtained from OpenStreetMap. The simulation software and data sets, which comprise gigabytes of spatio-temporal and temporal social network data, are made available to the research community.

The increasing democratization of server hardware with multi-core CPUs and large main memories has been one of the dominant hardware trends of the last decade. "Bare metal" servers with tens of CPU cores and over 100 gigabytes of main memory have been available for several years now. Recently, this large scale hardware has also been available via the cloud. Database systems, with their roots in uniprocessors and paucity of main memory, have unsurprisingly been found wanting on modern hardware. In addition to changes in hardware, database systems have had to contend with changing application requirements and deployment environments. Database systems have long provided applications with an interactive interface, in which an application can communicate with the database over several round-trips in the course of a single request. A large class of applications, however, does not require interactive interfaces, and is unwilling to pay the performance cost associated with overly flexible interfaces. Some of these applications have eschewed database systems altogether in favor of high-performance key-value stores. Finally, modern applications are increasingly deployed at ever increasing scales, often serving hundreds of thousands to millions of simultaneous clients. These large scale deployments are more prone to errors due to consistency issues in their underlying database systems. Ever since their inception, database systems have allowed applications to tradeoff consistency for performance, and often nudge applications towards weak consistency. When deployed at scale, weak consistency exposes latent consistency-related bugs, in the same way that failures are more likely to occur at scale. Nearly every widely deployed database system provides applications with weak consistency consistency by default, and its widespread use in practice significantly complicates application development, leading to latent Heisenbugs that are only exposed in production. This dissertation proposes and explores the use of deterministic execution to address these concerns. Database systems have traditionally been non-deterministic; given an input list of transactions, the final state of the database, which corresponds to some totally ordered execution of transactions, is dependent on non-deterministic factors such as thread scheduling decisions made by the operating system and failures. Deterministic execution, on the other hand, ensures that the database's final state is always determined by its input list of transactions; in other words, the input list of transactions is the same as the total order of transactions that determines the database's state. While non-deterministic database systems expend significant resources in determining valid total orders of transactions, we show that deterministic systems can exploit simple and low-cost up-front total ordering of transactions to execute and schedule transactions much more efficiently. We show that deterministic execution enables low-overhead, highly-parallel scheduling mechanisms, that can address the performance limitations of existing database systems on modern hardware. Deterministic database systems are designed based on the assumption that applications can submit their transactions in one-shot prepared transactions, instead of multiple round-trips. Finally, we attempt to understand the fundamental reason for the observed performance differences between various consistency levels in database systems, and based on this understanding, show that we can exploit deterministic execution to provide strong consistency at a cost that is competitive with that offered by weak consistency levels.

The U.S. Food and Drug Administration (FDA) has approved two digital pathology systems for primary diagnosis. These systems produce and consume whole slide images (WSIs) constructed from glass slides using advanced digital slide scanners. WSIs can greatly improve the work ow of pathologists through the development of novel image analytics software for automatic detection of cellular and morphological features and disease diagnosis using histopathology slides. However, the gigabyte size of a WSI poses a serious challenge for storage and retrieval of millions of WSIs. In this paper, we propose a system for scalable storage of WSIs and fast retrieval of image tiles using DRAM. A WSI is partitioned into tiles and sub-tiles using a combination of a space-filling curve, recursive partitioning, and Dewey numbering. They are then stored as a collection of key-value pairs in DRAM. During retrieval, a tile is fetched using key-value lookups from DRAM. Through performance evaluation on a 24-node cluster using 100 WSIs, we observed that, compared to Apache Spark, our system was three times faster to store the 100 WSIs and 1,000 times faster to access a single tile achieving millisecond latency. Such fast access to tiles is highly desirable when developing deep learning-based image analytics solutions on millions of WSIs.

In today’s epoch, mobile phones have become cosmopolitan. Functions that were hitherto considered inconceivable with a handheld device are now trivial for the quotidian user. The two factors that have led to such reformist developments are the exponential increase in processing power of mobile devices and ever-increasing data transfer rates. The second factor, namely the transfer of data at high speeds is eminently dependent upon the generation of telecommunication networks being used. The latest version of telecommunication networks is 5G. However, as speeds have increased, so have the requirements of the consumers. Data hungry applications demand that hundreds of gigabytes of data be transferred, with minimum latency. This, combined with a high amount of exposure to millimeter-wave radiation poses various health challenges. Among these health concerns, one of the biggest trepidations involves negative ocular effects. Our aim is to propose a technique that utilizes lower frequency millimeter waves (MMW) and decreases their negative effects on human eyes without compromising on the speed and latency that 5G communication systems are meant to deliver. The solution to speed, latency, and high radiation exposure combines lower bandwidths for data transmission while bringing data closer to the consumer. This can be made by the amalgamation of 5G with edge computing. Hence, this paper proposes how the lower bandwidths of 5G technology enhanced with edge computing to decrease latency and the amount of MMW exposure can give us the perfect data transfer mechanism.

Cloud storage is a service model where data is stored, managed, remotely backed up, and made available to users over a network (usually the Internet). Users generally pay per-consumption per month for their cloud data storage. Although the cost per gigabyte has been driven down drastically, cloud storage companies have introduced operating expenses that can make the service more costly than consumers who have been negotiating for. Therefore, the more room you use, the greater the price that you are liable to pay. The main objective of this project is to reduce the need to purchase additional storage area. Measures have been proposed to avoid duplicate files in the centralized cloud storage area. For this, secured auditing methodologies have been introduced. This will enhance the maintenance and management of files in the storage area. In addition, file classification techniques have been introduced wherein the users can upload the file into one of the following classes : private, public, sensitive and confidential. Files that are categorised as private and public are deleted periodically thereby saving storage space and the files that are confidential and sensitive are stored free of tampering unless the user wishes to delete them manually. At any instance, any file's class can be modified. This will help in reducing the need to buy extra storage space. Further, customised file tagging techniques are proposed where each file can be saved with three tags which aids efficient management of files. Through these approaches, files can be efficiently stored, accessed and managed in the centralised cloud storage area.

You’ve implemented a system, and you’ve created the best database structure you can architect. Your system is taking off, and you have hundreds or thousands of clients and many gigabytes of customer data on your server. Things couldn’t be better, right? Then you get a notification that something has gone wrong with your database. Somehow, a malicious entity has gained information about all of your customers because you were storing that information in your database in plain text. This nightmare scenario could be you, if you don’t take at least some minimal measures to prevent it.

Arms control treaties rely on extensive collection, processing and analysis of data to ascertain compliance with treaty obligations. Monitoring and evaluation relies on a broad range of technologies for sensing, communications and processing to collect, transmit and analyze the data necessary to detect non-compliance. The effectiveness of a treaty verification system depends on the ability of the information management system to collect and evaluate the information necessary to ascertain compliance with the treaty obligations. The information management systems for treaty verification have the same architecture because they are similar in concept and operation. They collect, transmit, store, process and analyze data. The effectiveness of each of these operations is a function of the capabilities of the associated technology. This chapter presents an overview of the technologies available for use in each of the elements of a treaty information management system. In sensing, these range from low frequency seismic sensors to multispectral and hyperspectral imaging. Data can be transmitted in the order of hundreds of Gigabits per second through fiber optic cables for terrestrial communications and tens of Megabits per second for satellite channels, while available technologies allow for storing petabytes of data and processing speeds of tens of Gigabytes per second. These capabilities are more than sufficient to collect the necessary data for use in sophisticated analysis algorithms for detecting treaty violations.

Large digital sky surveys are becoming the dominant source of data in astronomy. There are more than 100 terabytes of data in major archives, and that amount is growing rapidly. A typical sky survey archive has approximately 10 terabytes of image data and a billion detected sources (stars, galaxies, quasars, etc.), with hundreds of measured attributes per source. These surveys span the full range of wavelengths, radio through gamma ray, yet they are just a taste of the much larger datasets to come. Yearly advances in electronics bring new instruments that double the amount of data collected each year and lead to the exponential growth of information in astronomy. Thus, datasets that are orders of magnitude larger, more complex, and more homogeneous than in the past are on the horizon. In comparison, the size of the human genome is about 1 gigabyte and that of the Library of Congress is about 20 terabytes. Truly, astronomy has come to the big data era.

Deep Learning (DL) training platforms are built by interconnecting multiple DL accelerators (e.g., GPU/TPU) via fast, customized interconnects with 100s of gigabytes (GBs) of bandwidth. However, as we identify in this work, driving this bandwidth is quite challenging. This is because there is a pernicious balance between using the accelerator's compute and memory for both DL computations and communication. This work makes two key contributions. First, via real system measurements and detailed modeling, we provide an understanding of compute and memory bandwidth demands for DL compute and comms. Second, we propose a novel DL collective communication accelerator called Accelerator Collectives Engine (ACE) that sits alongside the compute and networking engines at the accelerator endpoint. ACE frees up the endpoint's compute and memory resources for DL compute, which in turn reduces the required memory BW by 3.5X on average to drive the same network BW compared to state-of-the-art baselines. For modern DL workloads and different network sizes, ACE, on average, increases the effective network bandwidth utilization by 1.44X (up to 2.67X), resulting in an average of 1.41X (up to 1.51X), 1.12X (up to 1.17X), and 1.13X (up to 1.19X) speedup in iteration time for ResNet-50, GNMT and DLRM when compared to the best baseline configuration, respectively.

Motion-activated cameras (camera traps) are increasingly used in ecological and management studies for remotely observing wildlife and have been regarded as among the most powerful tools for wildlife research. However, studies involving camera traps result in millions of images that need to be analyzed, typically by visually observing each image, in order to extract data that can be used in ecological analyses. We trained machine learning models using convolutional neural networks with the ResNet-18 architecture and 3,367,383 images to automatically classify wildlife species from camera trap images obtained from five states across the United States. We tested our model on an independent subset of images not seen during training from the United States and on an out-of-sample (or out-of-distribution in the machine learning literature) dataset of ungulate images from Canada. We also tested the ability of our model to distinguish empty images from those with animals in another out-of-sample dataset from Tanzania, containing a faunal community that was novel to the model. The trained model classified approximately 2,000 images per minute on a laptop computer with 16 gigabytes of RAM. The trained model achieved 98% accuracy at identifying species in the United States, the highest accuracy of such a model to date. Out-of-sample validation from Canada achieved 82% accuracy, and correctly identified 94% of images containing an animal in the dataset from Tanzania. We provide an R package (Machine Learning for Wildlife Image Classification; MLWIC) that allows the users to A) implement the trained model presented here and B) train their own model using classified images of wildlife from their studies. The use of machine learning to rapidly and accurately classify wildlife in camera trap images can facilitate non-invasive sampling designs in ecological studies by reducing the burden of manually analyzing images. We present an R package making these methods accessible to ecologists. We discuss the implications of this technology for ecology and considerations that should be addressed in future implementations of these methods.

The Compact Muon Solenoid (CMS) on the Large Hadron Collider (LHC) manage high volumes of data that currently exceeds 100PB across different sites. An important challenge of delivering data to experimenters in the CMS workflow is the data volume. An experiment data file has an average size of 2 Gigabytes, with file sizes ranging between 100 Megabytes and 20 Gigabytes. Also, a complete dataset comprises of multiple files, with the dataset files ranging from 2 Terabytes and 100 Terabytes in size. Providing fast access to datasets is an important enabler for data-intensive science research. In our work, we demonstrate a Information-Centric Networking (ICN) approach to providing fast in-network access to CMS datasets. To that end, we must first address the problem of how to store large CMS files in network caches closer to the end-users. We propose a software-defined, storage-aware routing mechanism using named data networking (NDN) to achieve this goal. Due to the inherent capacity limitations of the NDN router caches, we use software defined networking (SDN) to provide an intelligent and efficient solution for data distribution and routing across multiple NDN router caches. We demonstrate how software-defined control can be used for partitioning and distributing large CMS files based on NDN router cache-state knowledge. Further, SDN control will also configure the router forwarding strategy to retrieve CMS data from the network. Using our proposed architecture, we show that CMS dataset can be retrieved 28%–38% faster from the NDN routers caches compared to existing approaches. Lastly, we develop a prefetching mechanism to improve the transfer performance of files not available in the router's cache.

To eliminate redundant transfers over WAN links and improve network efficiency, middleboxes have been deployed at ingress/egress. These middleboxes can operate on individual packets and are application layer protocol transparent. They can identify and remove duplicated byte strings on the fly. However, with the increasing use of HTTPS, current redundancy elimination (RE) solution can no longer work without violating end-to-end privacy. In this paper, we present RE over encrypted traffic (REET), the first middlebox-based system that supports both intra-user and inter-user packet-level RE directly over encrypted traffic. REET realizes this by using a novel protocol with limited overhead and protects end users from honest-but-curious middleboxes. We implement REET and show its performance for both end users and middleboxes using several hundred gigabytes of network traffic traces collected from a large U.S. university.

A well Fifth generation (5G) mobile networks have been a common phrase in recent years. We have all heard this phrase and know its importance. By 2025, the number of devices based on the fifth generation of mobile networks will reach about 100 billion devices. By then, about 2.5 billion users are expected to consume more than a gigabyte of streamed data per month. 5G will play important roles in a variety of new areas, from smart homes and cars to smart cities, virtual reality and mobile augmented reality, and 4K video streaming. Bandwidth much higher than the fourth generation, more reliability and less latency are some of the features that distinguish this generation of mobile networks from previous generations. Clearly, at first glance, these features may seem very impressive and useful to a mobile network, but these features will pose serious challenges for operators and communications companies. All of these features will lead to considerable complexity. Managing this network, preventing errors, and minimizing latency are some of the challenges that the 5th generation of mobile networks will bring. Therefore, the use of artificial intelligence and machine learning is a good way to solve these challenges. in other say, in such a situation, proper management of the 5G network must be done using powerful tools such as artificial intelligence. Various researches in this field are currently being carried out. Research that enables automated management and servicing and reduces human error as much as possible. In this paper, we will review the artificial intelligence techniques used in communications networks. Creating a robust and efficient communications network using artificial intelligence techniques is a great incentive for future research. The importance of this issue is such that the sixth generation (6G) of cellular communications; There is a lot of emphasis on the use of artificial intelligence.

The tremendous boost in the next generation sequencing technologies and in the “omics” technologies resulted in the generation of hundreds of gigabytes of data per day. Nowadays, via integrating -omics data with other data types, such as imaging and electronic health record (EHR) data, panomics studies attempt to identify novel and potentially actionable biomarkers for personalized medicine applications. In this respect, for the accurate analysis of -omics data and EHR, there is a need to establish secure and robust pipelines that take the ethical aspects into consideration, regulate privacy and ownership issues, and data sharing. These days, blockchain technology has picked up significant attention in diverse fields, including genomics, since it offers a new solution for these problems from a different perspective. Blockchain is an immutable transaction ledger, which offers secure and distributed system without a central authority. Within the system, each transaction can be expressed with cryptographically signed blocks, and the verification of transactions is performed by the users of the network. In this review, firstly, we aim to highlight the challenges of EHR and genomic data sharing. Secondly, we attempt to answer “Why” or “Why not” the blockchain technology is suitable for genomics and healthcare applications in detail. Thirdly, we elucidate the general blockchain structure based on the Ethereum, which is a more suitable technology for the genomic data sharing platforms. Fourthly, we review current blockchain-based EHR and genomic data sharing platforms, evaluate the advantages and disadvantages of these applications, and classify these applications using different metrics. Finally, we conclude by discussing the open issues and introducing our suggestion on the topic. In summary, to facilitate the diagnosis, monitoring and therapy of diseases with the effective analysis of -omics data with other available data types, through this review, we put forward the possible implications of the blockchain technology to life sciences and healthcare.

The lack of large-scale open-source expert-labelled seismic datasets is one of the barriers to applying today’s AI techniques to automatic fault recognition tasks. The dataset present in this article consists of a large number of processed seismic images and their corresponding fault annotations. The processed seismic images, which are originally from a seismic survey called Thebe Gas Field in the Exmouth Plateau of the Carnarvan Basin on the NW shelf of Australia, are represented in Python Numpy format, which can be easily adopted by various AI models and will facilitate cooperation with researchers in the field of computer science. The corresponding fault annotations were firstly manually labelled by expert interpreters of faults from seismic data in order to investigate the structural style and associated evolution of the basin. Then the fault interpretation and seismic survey are processed and collected using Petrel software and Python programs separately. This dataset can help to train, validate, and evaluate the performance of different automatic fault recognition workflow.