Showing papers by "Manchester Metropolitan University published in 2021"

Review of deep learning: concepts, CNN architectures, challenges, applications, future directions

Abstract: In the last few years, the deep learning (DL) computing paradigm has been deemed the Gold Standard in the machine learning (ML) community. Moreover, it has gradually become the most widely used computational approach in the field of ML, thus achieving outstanding results on several complex cognitive tasks, matching or even beating those provided by human performance. One of the benefits of DL is the ability to learn massive amounts of data. The DL field has grown fast in the last few years and it has been extensively used to successfully address a wide range of traditional applications. More importantly, DL has outperformed well-known ML techniques in many domains, e.g., cybersecurity, natural language processing, bioinformatics, robotics and control, and medical information processing, among many others. Despite it has been contributed several works reviewing the State-of-the-Art on DL, all of them only tackled one aspect of the DL, which leads to an overall lack of knowledge about it. Therefore, in this contribution, we propose using a more holistic approach in order to provide a more suitable starting point from which to develop a full understanding of DL. Specifically, this review attempts to provide a more comprehensive survey of the most important aspects of DL and including those enhancements recently added to the field. In particular, this paper outlines the importance of DL, presents the types of DL techniques and networks. It then presents convolutional neural networks (CNNs) which the most utilized DL network type and describes the development of CNNs architectures together with their main features, e.g., starting with the AlexNet network and closing with the High-Resolution network (HR.Net). Finally, we further present the challenges and suggested solutions to help researchers understand the existing research gaps. It is followed by a list of the major DL applications. Computational tools including FPGA, GPU, and CPU are summarized along with a description of their influence on DL. The paper ends with the evolution matrix, benchmark datasets, and summary and conclusion.

CorrAUC: A Malicious Bot-IoT Traffic Detection Method in IoT Network Using Machine-Learning Techniques

Abstract: Identification of anomaly and malicious traffic in the Internet-of-Things (IoT) network is essential for the IoT security to keep eyes and block unwanted traffic flows in the IoT network. For this purpose, numerous machine-learning (ML) technique models are presented by many researchers to block malicious traffic flows in the IoT network. However, due to the inappropriate feature selection, several ML models prone misclassify mostly malicious traffic flows. Nevertheless, the significant problem still needs to be studied more in-depth that is how to select effective features for accurate malicious traffic detection in the IoT network. To address the problem, a new framework model is proposed. First, a novel feature selection metric approach named CorrAUC is proposed, and then based on CorrAUC, a new feature selection algorithm named CorrAUC is developed and designed, which is based on the wrapper technique to filter the features accurately and select effective features for the selected ML algorithm by using the area under the curve (AUC) metric. Then, we applied the integrated TOPSIS and Shannon entropy based on a bijective soft set to validate selected features for malicious traffic identification in the IoT network. We evaluate our proposed approach by using the Bot-IoT data set and four different ML algorithms. The experimental results analysis showed that our proposed method is efficient and can achieve >96% results on average.

A metaheuristic optimization approach for energy efficiency in the IoT networks

Abstract: © 2020 John Wiley & Sons, Ltd. Recently Internet of Things (IoT) is being used in several fields like smart city, agriculture, weather forecasting, smart grids, waste management, etc. Even though IoT has huge potential in several applications, there are some areas for improvement. In the current work, we have concentrated on minimizing the energy consumption of sensors in the IoT network that will lead to an increase in the network lifetime. In this work, to optimize the energy consumption, most appropriate Cluster Head (CH) is chosen in the IoT network. The proposed work makes use of a hybrid metaheuristic algorithm, namely, Whale Optimization Algorithm (WOA) with Simulated Annealing (SA). To select the optimal CH in the clusters of IoT network, several performance metrics such as the number of alive nodes, load, temperature, residual energy, cost function have been used. The proposed approach is then compared with several state-of-the-art optimization algorithms like Artificial Bee Colony algorithm, Genetic Algorithm, Adaptive Gravitational Search algorithm, WOA. The results prove the superiority of the proposed hybrid approach over existing approaches.

Energy-Efficient Random Access for LEO Satellite-Assisted 6G Internet of Remote Things

Abstract: Satellite communication system is expected to play a vital role for realizing various remote Internet-of-Things (IoT) applications in sixth-generation vision. Due to unique characteristics of satellite environment, one of the main challenges in this system is to accommodate massive random access (RA) requests of IoT devices while minimizing their energy consumptions. In this article, we focus on the reliable design and detection of RA preamble to effectively enhance the access efficiency in high-dynamic low-earth-orbit (LEO) scenarios. To avoid additional signaling overhead and detection process, a long preamble sequence is constructed by concatenating the conjugated and circularly shifted replicas of a single root Zadoff–Chu (ZC) sequence in RA procedure. Moreover, we propose a novel impulse-like timing metric based on length-alterable differential cross-correlation (LDCC), that is immune to carrier frequency offset (CFO) and capable of mitigating the impact of noise on timing estimation. Statistical analysis of the proposed metric reveals that increasing correlation length can obviously promote the output signal-to-noise power ratio, and the first-path detection threshold is independent of noise statistics. Simulation results in different LEO scenarios validate the robustness of the proposed method to severe channel distortion, and show that our method can achieve significant performance enhancement in terms of timing estimation accuracy, success probability of first access, and mean normalized access energy, compared with the existing RA methods.

Endothelial function in cardiovascular medicine: a consensus paper of the European Society of Cardiology Working Groups on Atherosclerosis and Vascular Biology, Aorta and Peripheral Vascular Diseases, Coronary Pathophysiology and Microcirculation, and Thrombosis.

Abstract: Endothelial cells (ECs) are sentinels of cardiovascular health. Their function is reduced by the presence of cardiovascular risk factors, and is regained once pathological stimuli are removed. In this European Society for Cardiology Position Paper, we describe endothelial dysfunction as a spectrum of phenotypic states and advocate further studies to determine the role of EC subtypes in cardiovascular disease. We conclude that there is no single ideal method for measurement of endothelial function. Techniques to measure coronary epicardial and micro-vascular function are well established but they are invasive, time-consuming, and expensive. Flow-mediated dilatation (FMD) of the brachial arteries provides a non-invasive alternative but is technically challenging and requires extensive training and standardization. We, therefore, propose that a consensus methodology for FMD is universally adopted to minimize technical variation between studies, and that reference FMD values are established for different populations of healthy individuals and patient groups. Newer techniques to measure endothelial function that are relatively easy to perform, such as finger plethysmography and the retinal flicker test, have the potential for increased clinical use provided a consensus is achieved on the measurement protocol used. We recommend further clinical studies to establish reference values for these techniques and to assess their ability to improve cardiovascular risk stratification. We advocate future studies to determine whether integration of endothelial function measurements with patient-specific epigenetic data and other biomarkers can enhance the stratification of patients for differential diagnosis, disease progression, and responses to therapy.

Energy-Aware Marine Predators Algorithm for Task Scheduling in IoT-Based Fog Computing Applications

Abstract: To improve the quality of service (QoS) needed by several applications areas, the Internet of Things (IoT) tasks are offloaded into the fog computing instead of the cloud. However, the availability of ongoing energy heads for fog computing servers is one of the constraints for IoT applications because transmitting the huge quantity of the data generated using IoT devices will produce network bandwidth overhead and slow down the responsive time of the statements analyzed. In this article, an energy-aware model basis on the marine predators algorithm (MPA) is proposed for tackling the task scheduling in fog computing (TSFC) to improve the QoSs required by users. In addition to the standard MPA, we proposed the other two versions. The first version is called modified MPA (MMPA), which will modify MPA to improve their exploitation capability by using the last updated positions instead of the last best one. The second one will improve MMPA by the ranking strategy based reinitialization and mutation toward the best, in addition to reinitializing, the half population randomly after a predefined number of iterations to get rid of local optima and mutated the last half toward the best-so-far solution. Accordingly, MPA is proposed to solve the continuous one, whereas the TSFC is considered a discrete one, so the normalization and scaling phase will be used to convert the standard MPA into a discrete one. The three versions are proposed with some other metaheuristic algorithms and genetic algorithms based on various performance metrics such as energy consumption, makespan, flow time, and carbon dioxide emission rate. The improved MMPA could outperform all the other algorithms and the other two versions.

Robust Spammer Detection Using Collaborative Neural Network in Internet-of-Things Applications

Abstract: Spamming is emerging as a key threat to the Internet of Things (IoT)-based social media applications. It will pose serious security threats to the IoT cyberspace. To this end, artificial intelligence-based detection and identification techniques have been widely investigated. The literature works on IoT cyberspace can be categorized into two categories: 1) behavior pattern-based approaches and 2) semantic pattern-based approaches. However, they are unable to effectively handle concealed, complicated, and changing spamming activities, especially in the highly uncertain environment of the IoT. To address this challenge, in this article, we exploit the collaborative awareness of both patterns, and propose a Collaborative neural network-based spammer detection mechanism (Co-Spam) in social media applications. In particular, it introduces multisource information fusion by collaboratively encoding long-term behavioral and semantic patterns. Hence, a more comprehensive representation of the feature space can be captured for further spammer detection. Empirically, we implement a series of experiments on two real-world data sets under different scenarios and parameter settings. The efficiency of the proposed Co-Spam is compared with five baselines with respect to several evaluation metrics. The experimental results indicate that the Co-Spam has an average performance improvement of approximately 5% compared to the baselines.

Annotated checklist of the birds of Brazil by the Brazilian Ornithological Records Committee—second edition

Abstract: An updated version of the checklist of birds of Brazil is presented, along with a summary of the changes approved by the Brazilian Ornithological Records Committee’s Taxonomy Subcommittee since the first edition, published in 2015. In total, 1971 bird species occurring in Brazil are supported by documentary evidence and are admitted to the Primary List, 4.3% more than in the previous edition. Eleven additional species are known only from undocumented records (Secondary List). For each species on the Primary List, status of occurrence in the country is provided and, in the case of polytypic species, the respective subspecies present in Brazilian territory are listed. Explanatory notes cover taxonomic changes, nomenclatural corrections, new occurrences, and other changes implemented since the last edition. Ninety species are added to the Primary List as a result of species descriptions, new occurrences, taxonomic splits, and transfers from the Secondary List due to the availability of documentation. In contrast, eight species are synonymized or assigned subspecific status and thus removed from the Primary List. In all, 293 species are endemic to Brazil, ranked third among the countries with the highest rate of bird endemism. The Brazilian avifauna currently consists of 1742 residents or breeding migrants, 126 seasonal non-breeding visitors, and 103 vagrants. The category of vagrants showed the greatest increase (56%) compared to the previous list, mainly due to new occurrences documented in recent years by citizen scientists. The list updates the diversity, systematics, taxonomy, scientific and vernacular nomenclature, and occurrence status of birds in Brazil.

Screen-printed electrodes: Transitioning the laboratory in-to-the field

Abstract: This short article overviews the use of screen-printed electrodes (SPEs) in the field of electroanalysis and compares their application against traditional laboratory based analytical techniques. Electroanalysis coupled with SPEs can offer low-cost, precise, sensitive, rapid, quantitative information and laboratory equivalent results. The combined use of SPEs and electroanalysis reduces the need of sample transportation and preparation to a centralised laboratory allowing experimentalists to perform the measurements where they are needed the most. We first introduce the basic concepts and principles of analytical techniques to the reader, with particular attention to electroanalysis, and then discuss the application of SPEs to common analytical targets such as food, environmental, forensics, cancer biomarkers and pathogenic monitoring and sensing.

Blockchain-enabled supply chain: analysis, challenges, and future directions

Abstract: Managing the integrity of products and processes in a multi-stakeholder supply chain environment is a significant challenge. Many current solutions suffer from data fragmentation, lack of reliable provenance, and diverse protocol regulations across multiple distributions and processes. Amongst other solutions, Blockchain has emerged as a leading technology, since it provides secure traceability and control, immutability, and trust creation among stakeholders in a low cost IT solution. Although Blockchain is making a significant impact in many areas, there are many impediments to its widespread adoption in supply chains. This article is the first survey of its kind, with detailed analysis of the challenges and future directions in Blockchain-enabled supply chains. We review the existing digitalization of the supply chain including the role of GS1 standards and technologies. Current use cases and startups in the field of Blockchain-enabled supply chains are reviewed and presented in tabulated form. Technical and non-technical challenges in the adoption of Blockchain for supply chain applications are critically analyzed, along with the suitability of various consensus algorithms for applications in the supply chain. The tools and technologies in the Blockchain ecosystem are depicted and analyzed. Some key areas as future research directions are also identified which must be addressed to realize mass adoption of Blockchain-based in supply chain traceability. Finally, we propose MOHBSChain, a novel framework for Blockchain-enabled supply chains.

COVID-19 lockdown: A perfect storm for older people's mental health.

Abstract: The management of lockdown presents a perfect storm for mental distress for older people by enforcing isolation and heightening perceptions of risk of death and illness. While gradual release from lockdown will maintain protection of those most at risk from covid‐19, older people will experience social isolation for the longest period as the over 75s carry the highest mortality risk (WHO, 2020). Isolation is strongly linked to depression, anxiety, and cognitive decline, and reduces resilience factors such as self‐worth, sense of purpose and feeling valued (Novotney, 2019). However, in addition to sustained isolation, governmental management of lockdown presents other challenges for older people.

How many bird and mammal extinctions has recent conservation action prevented

Abstract: Aichi Target 12 of the Convention on Biological Diversity (CBD) contains the aim to ‘prevent extinctions of known threatened species’. To measure the degree to which this was achieved, we used expert elicitation to estimate the number of bird and mammal species whose extinctions were prevented by conservation action in 1993–2020 (the lifetime of the CBD) and 2010–2020 (the timing of Aichi Target 12). We found that conservation action prevented 21–32 bird and 7–16 mammal extinctions since 1993, and 9–18 bird and two to seven mammal extinctions since 2010. Many remain highly threatened and may still become extinct. Considering that 10 bird and five mammal species did go extinct (or are strongly suspected to) since 1993, extinction rates would have been 2.9–4.2 times greater without conservation action. While policy commitments have fostered significant conservation achievements, future biodiversity action needs to be scaled up to avert additional extinctions.

Tourism and Hospitality industry resilience during the Covid-19 pandemic: Evidence from England

Abstract: The tourism and hospitality industries have been particularly impacted by the Covid-19 pandemic, with widespread closures and later re-opening times than other areas of economic activity. However, ...

Federated Deep Learning for Zero-Day Botnet Attack Detection in IoT Edge Devices

Abstract: Deep Learning (DL) has been widely proposed for botnet attack detection in Internet of Things (IoT) networks. However, the traditional Centralized DL (CDL) method cannot be used to detect previously unknown (zero-day) botnet attack without breaching the data privacy rights of the users. In this paper, we propose Federated Deep Learning (FDL) method for zero-day botnet attack detection to avoid data privacy leakage in IoT edge devices. In this method, an optimal Deep Neural Network (DNN) architecture is employed for network traffic classification. A model parameter server remotely coordinates the independent training of the DNN models in multiple IoT edge devices, while Federated Averaging (FedAvg) algorithm is used to aggregate local model updates. A global DNN model is produced after a number of communication rounds between the model parameter server and the IoT edge devices. Zero-day botnet attack scenarios in IoT edge devices is simulated with the Bot-IoT and N-BaIoT data sets. Experiment results show that FDL model: (a) detects zero-day botnet attacks with high classification performance; (b) guarantees data privacy and security; (c) has low communication overhead (d) requires low memory space for the storage of training data; and (e) has low network latency. Therefore, FDL method outperformed CDL, Localized DL, and Distributed DL methods in this application scenario.

Hybrid Deep Learning for Botnet Attack Detection in the Internet-of-Things Networks

Abstract: Deep learning (DL) is an efficient method for botnet attack detection. However, the volume of network traffic data and memory space required is usually large. It is, therefore, almost impossible to implement the DL method in memory-constrained Internet-of-Things (IoT) devices. In this article, we reduce the feature dimensionality of large-scale IoT network traffic data using the encoding phase of long short-term memory autoencoder (LAE). In order to classify network traffic samples correctly, we analyze the long-term inter-related changes in the low-dimensional feature set produced by LAE using deep bidirectional long short-term memory (BLSTM). Extensive experiments are performed with the BoT-IoT data set to validate the effectiveness of the proposed hybrid DL method. Results show that LAE significantly reduced the memory space required for large-scale network traffic data storage by 91.89%, and it outperformed state-of-the-art feature dimensionality reduction methods by 18.92–27.03%. Despite the significant reduction in feature size, the deep BLSTM model demonstrates robustness against model underfitting and overfitting. It also achieves good generalisation ability in binary and multiclass classification scenarios.

Impacts of COVID-19 and social isolation on academic staff and students at universities: a cross-sectional study.

Abstract: Background: “The impacts of the Coronavirus Disease 2019 (COVID-19) pandemic and the shutdown it triggered at universities across the world, led to a great degree of social isolation among university staff and students. The aim of this study was to identify the perceived consequences of this on staff and their work and on students and their studies at universities. Method: The study used a variety of methods, which involved an on-line survey on the influences of social isolation using a non-probability sampling. More specifically, two techniques were used, namely a convenience sampling (i.e. involving members of the academic community, which are easy to reach by the study team), supported by a snow ball sampling (recruiting respondents among acquaintances of the participants). A total of 711 questionnaires from 41 countries were received. Descriptive statistics were deployed to analyse trends and to identify socio-demographic differences. Inferential statistics were used to assess significant differences among the geographical regions, work areas and other socio-demographic factors related to impacts of social isolation of university staff and students. Results: The study reveals that 90% of the respondents have been affected by the shutdown and unable to perform normal work or studies at their institution for between 1 week to 2 months. While 70% of the respondents perceive negative impacts of COVID 19 on their work or studies, more than 60% of them value the additional time that they have had indoors with families and others. Conclusions: While the majority of the respondents agree that they suffered from the lack of social interaction and communication during the social distancing/isolation, there were significant differences in the reactions to the lockdowns between academic staff and students. There are also differences in the degree of influence of some of the problems, when compared across geographical regions. In addition to policy actions that may be deployed, further research on innovative methods of teaching and communication with students is needed in order to allow staff and students to better cope with social isolation in cases of new or recurring pandemics.

The educational response to Covid-19 across two countries: a critical examination of initial digital pedagogy adoption

Abstract: In spring 2020, K–12 schools adopted remote learning amidst the Covid-19 pandemic. Using activity theory, the authors examine the educational response to this global health crisis in the United Sta...

Dual Attention Multi-Instance Deep Learning for Alzheimer’s Disease Diagnosis With Structural MRI

Abstract: Structural magnetic resonance imaging (sMRI) is widely used for the brain neurological disease diagnosis, which could reflect the variations of brain. However, due to the local brain atrophy, only a few regions in sMRI scans have obvious structural changes, which are highly correlative with pathological features. Hence, the key challenge of sMRI-based brain disease diagnosis is to enhance the identification of discriminative features. To address this issue, we propose a dual attention multi-instance deep learning network (DA-MIDL) for the early diagnosis of Alzheimer’s disease (AD) and its prodromal stage mild cognitive impairment (MCI). Specifically, DA-MIDL consists of three primary components: 1) the Patch-Nets with spatial attention blocks for extracting discriminative features within each sMRI patch whilst enhancing the features of abnormally changed micro-structures in the cerebrum, 2) an attention multi-instance learning (MIL) pooling operation for balancing the relative contribution of each patch and yield a global different weighted representation for the whole brain structure, and 3) an attention-aware global classifier for further learning the integral features and making the AD-related classification decisions. Our proposed DA-MIDL model is evaluated on the baseline sMRI scans of 1689 subjects from two independent datasets (i.e., ADNI and AIBL). The experimental results show that our DA-MIDL model can identify discriminative pathological locations and achieve better classification performance in terms of accuracy and generalizability, compared with several state-of-the-art methods.

Evaluating the climate impact of aviation emission scenarios towards the Paris agreement including COVID-19 effects.

Abstract: Aviation is an important contributor to the global economy, satisfying society’s mobility needs. It contributes to climate change through CO2 and non-CO2 effects, including contrail-cirrus and ozone formation. There is currently significant interest in policies, regulations and research aiming to reduce aviation’s climate impact. Here we model the effect of these measures on global warming and perform a bottom-up analysis of potential technical improvements, challenging the assumptions of the targets for the sector with a number of scenarios up to 2100. We show that although the emissions targets for aviation are in line with the overall goals of the Paris Agreement, there is a high likelihood that the climate impact of aviation will not meet these goals. Our assessment includes feasible technological advancements and the availability of sustainable aviation fuels. This conclusion is robust for several COVID-19 recovery scenarios, including changes in travel behaviour.

A Fuzzy Detection System for Rumors through Explainable Adaptive Learning

Abstract: Nowadays, rumor spreading has gradually evolved into a kind of organized behaviors, accompanied with strong uncertainty and fuzziness However, existing fuzzy detection techniques for rumors focused their attention on supervised scenarios which require expert samples with labels for training Thus they are not able to well handle unsupervised scenarios where labels are unavailable To bridge such gap, this paper proposes a fuzzy detection system for rumors through explainable adaptive learning Specifically, its core is a graph embedding-based generative adversarial network (Graph-GAN) model First of all, it constructs fine-grained feature spaces via graph-level encoding Furthermore, it introduces continuous adversarial training between a generator and a discriminator for unsupervised decoding The two-stage scheme not only solves fuzzy rumor detection under unsupervised scenarios, but also improves robustness of the unsupervised training Empirically, a set of experiments are carried out based on three real-world datasets Compared with seven benchmark methods in terms of four metrics, the results of Graph-GAN reveal a proper performance which averagely exceeds baselines by 5% to 10%

COVID-19 Detection from Chest X-ray Images Using Feature Fusion and Deep Learning.

Abstract: Currently, COVID-19 is considered to be the most dangerous and deadly disease for the human body caused by the novel coronavirus. In December 2019, the coronavirus spread rapidly around the world, thought to be originated from Wuhan in China and is responsible for a large number of deaths. Earlier detection of the COVID-19 through accurate diagnosis, particularly for the cases with no obvious symptoms, may decrease the patient’s death rate. Chest X-ray images are primarily used for the diagnosis of this disease. This research has proposed a machine vision approach to detect COVID-19 from the chest X-ray images. The features extracted by the histogram-oriented gradient (HOG) and convolutional neural network (CNN) from X-ray images were fused to develop the classification model through training by CNN (VGGNet). Modified anisotropic diffusion filtering (MADF) technique was employed for better edge preservation and reduced noise from the images. A watershed segmentation algorithm was used in order to mark the significant fracture region in the input X-ray images. The testing stage considered generalized data for performance evaluation of the model. Cross-validation analysis revealed that a 5-fold strategy could successfully impair the overfitting problem. This proposed feature fusion using the deep learning technique assured a satisfactory performance in terms of identifying COVID-19 compared to the immediate, relevant works with a testing accuracy of 99.49%, specificity of 95.7% and sensitivity of 93.65%. When compared to other classification techniques, such as ANN, KNN, and SVM, the CNN technique used in this study showed better classification performance. K-fold cross-validation demonstrated that the proposed feature fusion technique (98.36%) provided higher accuracy than the individual feature extraction methods, such as HOG (87.34%) or CNN (93.64%).

The impact of Covid-19 on sport

Abstract: The Covid-19 pandemic has had an unprecedented impact on society, leading to a rapid closure of businesses, places of work, worship, social engagement, schools and universities. Sport is often seen...