Search

Availability of data and materials: The dataset used for supporting the conclusions of this paper is available from the public data repository at https://archive.ics.uci.edu/ml/index.php. The dataset is available at: https://archive.ics.uci.edu/ml/datasets/default+of+credit+card+clients. Copyright © The Author(s) 2021. With the rapid growth of consumer credit and the huge amount of financial data developing effective credit scoring models is very crucial. Researchers have developed complex credit scoring models using statistical and artificial intelligence (AI) techniques to help banks and financial institutions to support their financial decisions. Neural networks are considered as a mostly wide used technique in finance and business applications. Thus, the main aim of this paper is to help bank management in scoring credit card clients using machine learning by modelling and predicting the consumer behaviour with respect to two aspects: the probability of single and consecutive missed payments for credit card customers. The proposed model is based on the bidirectional Long-Short Term Memory (LSTM) model to give the probability of a missed payment during the next month for each customer. The model was trained on a real credit card dataset and the customer behavioural scores are analysed using classical measures such as accuracy, Area Under the Curve, Brier score, Kolmogorov–Smirnov test, and H-measure. Calibration analysis of the LSTM model scores showed that they can be considered as probabilities of missed payments. The LSTM model was compared to four traditional machine learning algorithms: support vector machine, random forest, multi-layer perceptron neural network, and logistic regression. Experimental results show that, compared with traditional methods, the consumer credit scoring method based on the LSTM neural network has significantly improved consumer credit scoring. Office of Research, Zayed University under Grant Number R20053.

In this paper, a new model known as YOLO-v5 is initiated to detect defects in PCB. In the past many models and different approaches have been implemented in the quality inspection for detection of defect in PCBs. This algorithm is specifically selected due to its efficiency, accuracy and speed. It is well known that the traditional YOLO models (YOLO, YOLO-v2, YOLO-v3, YOLO-v4 and Tiny-YOLO-v2) are the state-of-the-art in artificial intelligence industry. In electronics industry, the PCB is the core and the most basic component of any electronic product. PCB is almost used in each and every electronic product that we use in our daily life not only for commercial purposes, but also used in sensitive applications such defense and space exploration. These PCB should be inspected and quality checked to detect any kind of defects during the manufacturing process. Most of the electronic industries are focused on the quality of their product, a small error during manufacture or quality inspection of the electronic products such as PCB leads to a catastrophic end. Therefore, there is a huge revolution going on in the manufacturing industry where the object detection method like YOLO-v5 is a game changer for many industries such as electronic industries. Ministry of Science and Technology, Taiwan

© Copyright 2021, A. O. Nyanteh et al. Software-Defined Networking (SDN) is an innovative technology which provides a programmable network control which is decoupled from the physical infrastructure. Network Virtualization (NV) is the phenomenon where a given physical network infrastructure and its resources are abstracted to create multiple logical virtual network slices of the underlying substrate. NV enables independent virtual networks to co-exist on one or more shared physical network infrastructure. Edge computing makes use of the edge resources in close proximity to end-users to reduce service delay and the network traffic volume in the end-to-end networks. Similarly, network slicing which is a key enabling technology for 5G networks is designed to support different services from different platforms at different scales enables sharing of physical network infrastructure on many different virtual network layers. These innovative technologies and strategies have gained significant attention from both academia and industry as they have the potential to maximize network resource utilization and optimize end–to–end network service delivery in 5G solutions deployment. To enable continuous simulation and development of applicable 5G networking concepts using these technologies, there is a need for an accessible and easy-to-learn testbed which is able to efficiently measure the performance of physical and virtual network capacities, provisioning approaches and management of multiple architectural models using large-scale network slicing configurations in a repeatable and controllable manner. These tools and toolkits provide scalable, lightweight and controlled cloud simulation environments necessary to analyse network traffic flows, allocation capacities and policies and the behaviour of multiple heterogeneous networks at an extremely low cost as compared to the huge financial commitments involved in conducting similar experiments in a real-life event. Existing solutions do not support Network Slicing and end- to -end heterogenous network automation which are key enablers of 5G network implementation. Hence in this paper, the CloudSimHypervisor framework is developed in this based on CloudSimSDN-NFV. The complete architecture and features of the CloudSimHypervisor framework and some used cases are presented in this paper. We validate the CloudSimHypervisor framework with two use case experiments in the cloud computing environment: Joint compute and network resource utilization and network traffic prioritization. Results from these experiments display the efficiency of the CloudSimHypervisor in estimating and measuring processing speed, transmission speed, compute and network usage efficiency and energy consumption.

According to a recently conducted survey on surgical complication mortality rate, 47% of such cases are due to anesthetics overdose. This indicates that there is an urgent need to moderate the level of anesthesia. Recently deep learning (DL) methods have played a major role in estimating the depth of Anesthesia (DOA) of patients and has played an essential role in control anesthesia overdose. In this paper, Electroencephalography (EEG) signals have been used for the prediction of DOA. EEG signals are very complex signals which may require months of training and advanced signal processing techniques. It is a point of debate whether DL methods are an improvement over the already existing traditional EEG signal processing approaches. One of the DL algorithms is Convolutional neural network (CNN) which is very popular algorithm for object recognition and is widely growing its applications in processing hierarchy in the human visual system. In this paper, various decomposition methods have been used for extracting the features EEG signal. After acquiring the necessary signals values in image format, several CNN models have been deployed for classification of DOA depending upon their Bispectral Index (BIS) and the signal quality index (SQI). The EEG signals were converted into the frequency domain using and Empirical Mode Decomposition (EMD), and Ensemble Empirical Mode Decomposition (EEMD). However, because of the inter mode mixing observed in EMD method; EEMD have been utilized for this study. The developed CNN models were used to predict the DOA based on the EEG spectrum images without the use of handcrafted features which provides intuitive mapping with high efficiency and reliability. The best trained model gives an accuracy of 83.2%. Hence, this provides further scope and research which can be carried out in the domain of visual mapping of DOA using EEG signals and DL methods. Ministry of Science and Technology, Taiwan

Parkinson’s disease (PD) is a type of neurodegenerative diseases. PD influences gait in many aspects: reduced gait speed and step length, increased axial rigidity, and impaired rhythmicity. Gait-related data used in this study are from PhysioNet. Twenty-one PD patients and five healthy controls (CO) were sorted into four groups: PD without task (PDw), PD with dual task (PDd), control without task (COw), and control with dual task (COd). Since dual task actions are attention demanding, either gait or cognitive function may be affected. To quantify the used walking data, eight pressure sensors installed in each insole are used to measure the vertical ground reaction force. Thus, quantitative measurement analysis is performed utilizing multiscale entropy (MSE) and complexity index (CI) to analyze and differentiate between the ground reaction force of the four different groups. Results show that the CI of patients with PD is higher than that of CO and 11 of the sensor signals are statistically significant (p &lt, 0.05). The COd group has larger CI values at the beginning (p = 0.021) but they get lower at the end of the test (p = 0.000) compared to that in the COw group. The end-of-test CI for the PDw group is lower in one of the feet sensor signals, and in the right total ground reaction force compared to the PDd group counterparts. In conclusion, when people start to adjust their gait due to pathology or stress, CI may increase first and reach a peak, but it decreases afterward when stress or pathology is further increased.

Autonomous Vehicles (AVs) have the potential to solve many traffic problems, such as accidents, congestion and pollution. However, there are still challenges to overcome, for instance, AVs need to accurately perceive their environment to safely navigate in busy urban scenarios. The aim of this paper is to review recent articles on computer vision techniques that can be used to build an AV perception system. AV perception systems need to accurately detect non-static objects and predict their behaviour, as well as to detect static objects and recognise the information they are providing. This paper, in particular, focuses on the computer vision techniques used to detect pedestrians and vehicles. There have been many papers and reviews on pedestrians and vehicles detection so far. However, most of the past papers only reviewed pedestrian or vehicle detection separately. This review aims to present an overview of the AV systems in general, and then review and investigate several detection computer vision techniques for pedestrians and vehicles. The review concludes that both traditional and Deep Learning (DL) techniques have been used for pedestrian and vehicle detection; however, DL techniques have shown the best results. Although good detection results have been achieved for pedestrians and vehicles, the current algorithms still struggle to detect small, occluded, and truncated objects. In addition, there is limited research on how to improve detection performance in difficult light and weather conditions. Most of the algorithms have been tested on well-recognised datasets such as Caltech and KITTI; however, these datasets have their own limitations. Therefore, this paper recommends that future works should be implemented on more new challenging datasets, such as PIE and BDD100K. EPSRC DTP PhD studentship

Currently, no device is available for the rapid screening of ventilator-associated pneumonia (VAP) at an early stage. Accordingly, we propose the design of an offline gas detection system to monitor and detect metabolites of pneumonia at an early stage. An electronic nose (e-nose) with 28 metal oxide semiconductor gas sensors was developed for predicting the presence of infection after patients have been intubated in the intensive care unit. The effectiveness of VAP identification was verified using clinical data. A total of 40 patients were included in this study, of whom 20 were infected with Pseudomonas aeruginosa and the remaining were uninfected. The results revealed that good accuracy rates of 0.9208% ± 0.0302% and 0.8547% ± 0.0214% were achieved by support vector machine and artificial neural network models, respectively. This study provides a simple, low-cost solution for the rapid screening of VAP at an early stage.

Data Availability Statement: This study utilizes the publicly available dataset, from https://physionet.org, accessed on 14 January 2021. Copyright: © 2021 by the authors. This study evaluates cardiovascular and cerebral hemodynamics systems by only using non-invasive electrocardiography (ECG) signals. The Massachusetts General Hospital/Marquette Foundation (MGH/MF) and Cerebral Hemodynamic Autoregulatory Information System Database (CHARIS DB) from the PhysioNet database are used for cardiovascular and cerebral hemodynamics, respectively. For cardiovascular hemodynamics, the ECG is used for generating the arterial blood pressure (ABP), central venous pressure (CVP), and pulmonary arterial pressure (PAP). Meanwhile, for cerebral hemodynamics, the ECG is utilized for the intracranial pressure (ICP) generator. A deep convolutional autoencoder system is applied for this study. The cross-validation method with Pearson's linear correlation (R), root mean squared error (RMSE), and mean absolute error (MAE) are measured for the evaluations. Initially, the ECG is used to generate the cardiovascular waveform. For the ABP system-the systolic blood pressure (SBP) and diastolic blood pressures (DBP)-the R evaluations are 0.894 ± 0.004 and 0.881 ± 0.005, respectively. The MAE evaluations for SBP and DBP are, respectively, 6.645 ± 0.353 mmHg and 3.210 ± 0.104 mmHg. Furthermore, for the PAP system-the systolic and diastolic pressures-the R evaluations are 0.864 ± 0.003 mmHg and 0.817 ± 0.006 mmHg, respectively. The MAE evaluations for systolic and diastolic pressures are, respectively, 3.847 ± 0.136 mmHg and 2.964 ± 0.181 mmHg. Meanwhile, the mean CVP evaluations are 0.916 ± 0.001, 2.220 ± 0.039 mmHg, and 1.329 ± 0.036 mmHg, respectively, for R, RMSE, and MAE. For the mean ICP evaluation in cerebral hemodynamics, the R and MAE evaluations are 0.914 ± 0.003 and 2.404 ± 0.043 mmHg, respectively. This study, as a proof of concept, concludes that the non-invasive cardiovascular and cerebral hemodynamics systems can be potentially investigated by only using the ECG signal. Funding: This research received no external funding.

The connections of roof-top photovoltaic (PV) systems with low voltage distribution networks have been raised dramatically in the last years due to decreasing in the capital cost and investing in the operating cost. However, the power generation of this source is variable owing to the solar irradiance dependency. This variation in the PV generation can cause a Total Harmonic Distortion (THD) into the low voltage distribution network. In this paper, the effect of roof-top PV system installations on the total harmonic distortion is assessed under different weather conditions. Regarding this, the voltage and current of an on-grid PV system composited at the roof-top of Brunel university London campus are measured and analysed for sunny and cloudy days. Furthermore, this installed PV system is developed based on MATLAB/Simulink. The results prove that the solar irradiance of the surround PV system has a significant impact on the THD-current when compared with the THD-voltage.

Review Article Copyright © 2021 Anthony Bawa et al. Gait and posture studies have gained much prominence among researchers and have attracted the interest of clinicians. The ability to detect gait abnormality and posture disorder plays a crucial role in the diagnosis and treatment of some diseases. Microsoft Kinect is presented as a noninvasive sensor essential for medical diagnostic and therapeutic purposes. There are currently no relevant studies that attempt to summarise the existing literature on gait and posture abnormalities using Kinect technology. The purpose of this study is to critically evaluate the existing research on gait and posture abnormalities using the Kinect sensor as the main diagnostic tool. Our studies search identified 458 for gait abnormality, 283 for posture disorder of which 26 studies were included for gait abnormality, and 13 for posture. The results indicate that Kinect sensor is a useful tool for the assessment of kinematic features. In conclusion, Microsoft Kinect sensor is presented as a useful tool for gait abnormality, postural disorder analysis, and physiotherapy. It can also help track the progress of patients who are undergoing rehabilitation.

© 2021 by the authors. Pain is a subjective feeling; it is a sensation that every human being must have experienced all their life. Yet, its mechanism and the way to immune to it is still a question to be answered. This review presents the mechanism and correlation of pain and stress, their assessment and detection approach with medical devices and wearable sensors. Various physiological signals (i.e., heart activity, brain activity, muscle activity, Electrodermal activity, respiratory, blood volume pulse, skin tempera-ture) and behavioral signals are organized for wearables sensors detection. By reviewing the wearable sensors used in the healthcare domain, we hope to find a way for wearable healthcare-monitoring system to be applied on pain and stress detection. Since pain leads to multiple consequences or symptoms such as muscle tension and depression that are stress related, there is a chance to find a new approach for chronic pain detection using daily life sensors or de-vices. Then by integrating modern computing techniques, there is a chance to handle pain and stress management issue. Ministry of Science and Technology (MOST) of Taiwan (grant number: MOST 107-2221-E-155-009-MY2).

Cyber warfare is undoubtedly inevitable, since this can be done with low costs, minimal equipment, and it can be performed by attribution which means adversaries are less likely to be counter attacked. Industrial Control Systems (ICS) are a system which was built over thirty years ago, and a system without security in-mind. Therefore, these systems are vulnerable, however they should be protected diligently as these systems are critical to the economic growth, public healthcare and welfare. In this paper, we will look to critically appraise cyber warfare and its effects against ICS, along with robust recommendations with the intention of enhancing national security.

Magnetic resonance (MR) image segmentation is one of the most robust MR based attenuation correction methods which have been adopted in clinical routine for positron emission tomography (PET) quantification. However, the segmentation of the brain into different tissue classes is a challenging process due to the similarity between bone and air signal intensity values. The aim of this work is to study the feasibility of deep learning to improve the brain segmentation with the application of data augmentation. A deep convolutional auto encoder network is applied to segment the brain into three tissue classes: air, soft tissue, and bone. The dice similarity coefficients of air, soft tissue, and bone tissues are 0.96 ± 0.01, 0.86 ± 0.02, and 0.63 ± 0.06 respectively. Despite the small datasets used in this work, the results are promising and show the feasibility of deep learning with data augmentation to perform accurate segmentation.

Initial big data analytics (BDA) implementation regarding data mining mechanisms in the audit process, remain un-der-researched and hence, more investigation is needed in developing classification prediction models utilising data mining analytics to identify appropriate auditing opinion. Empirical results show that the decision tree technique provides a balanced model for choosing the right audit opinion, with it achieving higher evaluation rates for two datasets (for the years 2018 and 2019) than Naive Bayes Networks (NBN), Linear Discriminant Analysis (LDA), and Logistic Regression (LR). In addition, for both datasets the LR model achieved higher evaluation rates than NBN and LDA. Moreover, the evaluation results for NBN exhibited a gap between 2019 and 2018 datasets, which means that it is an unbalanced model and hence, is not a good predictor of audit opinion. This paper is a pioneering effort utilising the sample of a company that have big data, based on financial and non-financial data to develop a model able to predict the correct audit opinion.

Magnetic resonance (MR) image segmentation is a robust technique used for PET attenuation correction. However, the segmentation of the brain into different tissue classes is a challenging task because of the similarity of pixel intensity values. The objective of this work is to propose a deep learning network to segment T1-weighted MR images of a dataset consists of 50 patients. Additionally, transfer learning is applied to segment another MR image protocol which is T2-weighted. The pretrained network with T1-weighted images is finetuned then tested with a dataset of 14 patients only. The Dice coefficients of air, soft tissue, and bone classes for T1-weighted MR images are 0.98, 0.92, and 0.79 respectively. The results of transfer learning show the feasibility of finetuning a deep network trained with T1-weighted images to segment T2-weighted images.

Cardiovascular diseases affect approximately 50 million people worldwide; thus, heart disease prevention is one of the most important tasks of any health care system. Despite the high popularity electrocardiography, superior automatic electrocardiography (ECG) signal analysis methods are required. The aim of this research was to design a new deep learning method for effectively classifying arrhythmia by using 2-second segments of 2D recurrence plot images of ECG signals. In the first stage, the noise and ventricular fibrillation (VF) categories were distinguished. In the second stage, the atrial fibrillation (AF), normal, premature AF, and premature VF categories were distinguished. Models were trained and tested using ECG databases publicly available at the website of PhysioNet. The MIT-BIH Arrhythmia Database, Creighton University Ventricular Tachyarrhythmia Database, MIT-BIH Atrial Fibrillation Database, and MIT-BIH Malignant Ventricular Ectopy Database were used to compare six types of arrhythmia. Testing accuracies of up to 95.3 % ± 1.27 % and 98.41 % ± 0.11 % were achieved for arrhythmia detection in the first and second stage, respectively, after five-fold cross-validation. In conclusion, this study provides clinicians with an advanced methodology for detecting and discriminating between different arrhythmia types.

In wireless networks, huge throughput, minimum delay and optimum bandwidth efficiency are considered as the prime concerns when mobility has existence. In sensor networks, when nodes change position, the topology of the network can also amend rapidly, hence such networks seek for a responsive routing procedures that finds valid routes directly as the topology changes and old routes turned into broken. The restricted capacity of the network channel calls for efficient routing procedures that do not drive the network into a congested status when new routes are implemented. In this paper we strive to present a focus review and analysis for three fundamental routing categories in sensor networks based on the network structure, specifically data-centric protocols, hierarchical protocols and location based protocols. Hierarchical protocols are applied to achieve energy efficient routing, data centric protocols are inquiry based and concentrate on eliminating non essential transmissions whereas location based procedures need position coordinates in order that optimal path can be established without utilizing flooding scheme. This review provides a sprightly and structured introduction to the most important network structure routing protocols supported by illustrative layouts.

© 2020 by the authors. In this study, a deep learning algorithm based on the you-only-look-once (YOLO) approach is proposed for the quality inspection of printed circuit boards (PCBs). The high accuracy and efficiency of deep learning algorithms has resulted in their increased adoption in every field. Similarly, accurate detection of defects in PCBs by using deep learning algorithms, such as convolutional neural networks (CNNs), has garnered considerable attention. In the proposed method, highly skilled quality inspection engineers first use an interface to record and label defective PCBs. The data are then used to train a YOLO/CNN model to detect defects in PCBs. In this study, 11,000 images and a network of 24 convolutional layers and 2 fully connected layers were used. The proposed model achieved a defect detection accuracy of 98.79% in PCBs with a batch size of 32. Ministry of Science and Technology of Taiwan; Boardtek Electronics Corporation, Taiwan

The final authenticated version is available online at https://doi.org/10.1007/978-3-030-55180-3_11. © 2021, Springer Nature Switzerland AG. Despite the interest of developed countries in relying on artificial intelligence in the performance of their work, developing countries are at the beginning of this path. The researcher has noted that the use of modern technologies, especially artificial intelligence, did not give the necessary attention in government business and many private sector companies, so it was necessary to draw attention to the importance of using it to develop all businesses. This study focuses on clarifying the importance of using artificial intelligence (AI) technology in the process of evaluating the employee’s performance to increase the effectiveness of performance appraisal for all organisational levels to make strategic decisions which affect the objectives of the organisation. This study has used a questionnaire that looked at participants’ views and attitudes towards technology and its usefulness in the appraisal process. Knowing the opinions of employees and their supervisors will enable to demonstrate the need to use AI to raise the efficiency of the performance appraisal system and develop its processes for the employee’s benefits and the general interest at the organisation. Most of the answers were between agree and strongly agree, and on the other side, there was a very small percentage that did not believe the importance of using artificial intelligence in works, and this may be due to their feeling of resisting change and the desire to work a way they used to. This paper begins by providing demographic details and background information on the participants, followed by an examination of the reliability of the extracted factors/metrics. The extracted factors will then be described individually using general descriptive statistics (such as repetition percentages) to look at their respective elements and see the level of the agreement created by the participant. Following this descriptive part of the analysis, this paper seeks to conduct an in-depth analysis of the results using various inferential statistics to test group differences using demographic and background details of the participants (e.g. age, gender, education). This study investigates the impact of the use of AI on the performance appraisal model, and to examine the effect of AI on the technology adoption.

Recent emerging hybrid technology of positron emission tomography/magnetic resonance (PET/MR) imaging has generated a great need for an accurate MR image-based PET attenuation correction. MR image segmentation, as a robust and simple method for PET attenuation correction, has been clinically adopted in commercial PET/MR scanners. The general approach in this method is to segment the MR image into different tissue types, each assigned an attenuation constant as in an X-ray CT image. Machine learning techniques such as clustering, classification and deep networks are extensively used for brain MR image segmentation. However, only limited work has been reported on using deep learning in brain PET attenuation correction. In addition, there is a lack of clinical evaluation of machine learning methods in this application. The aim of this review is to study the use of machine learning methods for MR image segmentation and its application in attenuation correction for PET brain imaging. Furthermore, challenges and future opportunities in MR image-based PET attenuation correction are discussed.

One of the most challenging predictive data analysis efforts is accurate prediction of depth of anesthesia (DOA) indicators which has attracted a growing attention since it provides patients a safe surgical environment in case of secondary damage caused by intraoperative awareness or brain injury. However, many researchers put heavily handcraft feature extraction or carefully tailored feature engineering to each patient to achieve very high sensitivity and low false prediction rate for a particular dataset. This limits the benefit of the proposed approaches if a different dataset is used. Recently, representations learned using deep convolutional neural network (CNN) for object recognition are becoming widely used model of the processing hierarchy in the human visual system. The correspondence between models and brain signals that holds the acquired activity at high temporal resolution has been explored less exhaustively. In this paper, deep learning CNN with a range of different architectures, is designed for identifying related activities from raw electroencephalography (EEG). Specifically, an improved short-time Fourier transform (STFT) is used to stand for the time-frequency information after extracting the spectral images of the original EEG as input to CNN. Then CNN models are designed and trained to predict the DOA levels from EEG spectrum without handcrafted features, which presents an intuitive mapping process with high efficiency and reliability. As a result, the best trained CNN model achieved an accuracy of 93.50%, interpreted as CNN’s deep learning to approximate the DOA by senior anesthesiologists, which highlights the potential of deep CNN combined with advanced visualization techniques for EEG-based brain mapping. This research was financially supported by Lenovo Technology B.V. Taiwan Branch. Also, it was supported by National Chung-Shan Institute of Science & Technology in Taiwan (Grant nos. CSIST-095-V301 and CSIST-095-V302) and National Natural Science Foundation of China (Grant no. 51475342).

Improving the energy efficiency of Heating, Ventilation, and Air Conditioning (HVAC) systems is crucial to reduce buildings’ energy costs and their carbon footprint. HVAC systems are complex, large-scale structures with pure lag time and high thermal inertia. Although traditionally, physical-based methods have been used to model, control and optimise them, data-driven approaches have demonstrated to be more application relevant, easier to compute and better suited to handle nonlinearities. Based only on measured or estimated data, data-driven approaches are highly dependent on the quality of the used data. In recent years, the advances in Information and Communication Technology (ICT), decreasing hardware cost, and improving data accessibility, have allowed the collection and storage of a large amount of high-quality building-related data, allowing the development of more accurate and robust data-driven approaches, making them gain great popularity in HVAC applications. In this paper, a Systematic Literature Review (SLR) based on a database search is conducted to give an in-depth insight into the major challenges regarding modelling, controlling and optimising HVAC systems, making the especial focus on the capability of data-driven models to improve their energy performance while keeping the users’ comfort. The main results of the SLR highlight the importance of taking users’ needs into account when modelling, controlling and optimising HVAC systems to avoid their underutilisation. In particular, the increasing tendency to include users’ feedback into Model Predictive Control (MPC) loops and use easy-to-access technologies, such as WiFi and Smartphone Applications (Apps), to acquire users’ information suggests promising future research horizons.

Purpose The purpose of this study is to propose an objective and efficient method for assessing credit risk by introducing and investigating to a greater extent the applicability of credit scoring models in the Jordanian banks and to what range they can be used to achieve their strategic and business objectives. Design/methodology/approach The research methodology comprises two phases. The first phase is the model development. Three modeling techniques are used to build the scoring models, namely, logistic regression (LR), artificial neural network (NN) and support vector machine (SVM), and the best performing model is selected for next stage. The second phase is two-fold: linking the credit expert knowledge in a way that can enhance the outcomes of the scoring model and a profitability test to explore if the selected model is efficient in meeting banks’ strategic and business objectives. Findings The findings showed that LR model outperformed both ANN and SVM across various performance indicators. The LR model also fits best with achieving the bank’s strategic and business objectives. Originality/value To the best of the authors’ knowledge, this study is the first that applied several modeling and classification techniques for Jordanian banks and calibrated the best model in terms of its strategic and business objectives. Furthermore, credit experts’ knowledge was engaged with the scoring model to determine its efficiency and reliability against the sole use of an automated scoring model in the hope to encourage the application of credit scoring models as an advisory tool for credit decisions.

Phase-rectified signal averaging (PRSA) has been known to be a useful method to detect periodicities in non-stationary biological signals. Determination of quasi-periodicities in electroencephalogram (EEG) is a candidate for quantifying the changes of depth of anesthesia (DOA). In this paper, DOA monitoring capacity of periodicities detected using PRSA were quantified by assessing EEG signals collected from 56 patients during surgery. The method is compared to sample entropy (SampEn), detrended fluctuation analysis (DFA) and permutation entropy (PE). The performance of quasi-periodicities defined by acceleration capacity (AC) and deceleration capacity (DC) was tested using the area under the receiver operating characteristic curve (AUC) and Pearson correlation coefficient. During the surgery, a significant difference (p < 0.05) in the quasi-periodicities was observed among three different stages under general anesthesia. There is a larger mean AUC and correlation coefficient of quasi-periodicities compared to SampEn, DFA and PE using expert assessment of conscious level (EACL) and bispectral index (BIS) as the gold standard, respectively. Quasi-periodicities detected using PRSA in EEG signals are powerful monitor of DOA and perform more accurate and robust results compared to SampEn, DFA and PE. The results do provide a valuable reference to researchers in the filed of clinical applications. 10.13039/501100003711-Center for Dynamical Biomarkers and Translational Medicine, National Central University, Taiwan, which is sponsored by the Ministry of Science and Technology; 10.13039/501100001809-National Natural Science Foundation of China

The proposal of this study is a new nonlinear autoregressive moving average, NARMA-L2 controller, which is based on an adaptive neuro-fuzzy inference system, ANFIS architecture. The new control configuration employs Sugeno-type fuzzy inference system FIS submodels to map input characteristics to the output of a dynamic and nonlinear system. The commonly used learning algorithm, which is called a hybrid method (Backpropagation and Least Square Error), has been carried out as well as particle swarm optimisation (PSO) approach, in order to select the optimal parameters of the ANFIS submodels. Once the system has been modelled efficiently and accurately, the proposed controller is designed by rearranging the generalised submodels. The controller performance is evaluated by simulations conducted on a binary distillation column, which is characterised by a nonlinear and dynamic behaviour. The obtained results show that the PSO-ANFIS based NARMA-L2 achieved more efficient modelling and control performances when compared with other controllers. These controllers include ANN-based NARMA-L2, (PD, PI and PID like) fuzzy-tuned by GA and PSO and traditional PID, which are also implemented to the column for comparison. Stability and robustness of the proposed controller regarding system inputs variance have also been tested by applying asynchronous setpoints of both inputs of the process.

One of the ways of increasing recognition ability in classification problem is removing outlier entries as well as redundant and unnecessary features from training set. Filtering and feature selection can have large impact on classifier accuracy and area under the curve (AUC), as noisy data can confuse classifier and lead it to catch wrong patterns in training data. The common approach in data filtering is using proximity graphs. However, the problem of the optimal filtering parameters selection is still insufficiently researched. In this paper filtering procedure based on k-nearest neighbours proximity graph was used. Filtering parameters selection was adopted as the solution of outlier minimization problem: k-NN proximity graph, power of distance and threshold parameters are selected in order to minimize outlier percentage in training data. Then performance of six commonly used classifiers (Logistic Regression, Naïve Bayes, Neural Network, Random Forest, Support Vector Machine and Decision Tree) and one heterogeneous classifiers combiner (DES-LA) are compared with and without filtering. Dynamic ensemble selection (DES) systems work by estimating the level of competence of each classifier from a pool of classifiers. Only the most competent ones are selected to classify a given test sample. This is achieved by defining a criterion to measure the level of competence of base classifiers, such as, its accuracy in local regions of the feature space around the query instance. In our case the combiner is based on the local accuracy of single classifiers and its output is a linear combination of single classifiers ranking. As results of filtering, accuracy of DES-LA combiner shows big increase for low-accuracy datasets. But filtering doesn’t have sufficient impact on DES-LA performance while working with high-accuracy datasets. The results are discussed, and classifiers, which performance was highly affected by pre-processing filtering step, are defined. The main contribution of the paper is introducing modifications to the DES-LA combiner, as well as comparative analysis of filtering impact on the classifiers of various type. Testing the filtering algorithm on real case dataset (Taiwan default credit card dataset) confirmed the efficiency of automatic filtering approach.

Marketing is a process of creating, capturing, and exchanging ‘value' for the mutual benefits of marketers, customers, intermediaries, and other stakeholders. Such a transaction requires trust as it might be facing a range of online cyber risks. Modern cybercrimes have exponentially grown over the last decade. Ransomware is one of the types of malware which is the result of a sophisticated attempt to compromise the modern computer systems. The businesses, governments, and large corporations are investing heavily to combat this cyber threat against their critical infrastructure. New technological shifts help to improve marketing and business productivity and keep the company's global competitiveness in an overflowing competitive market. However, the businesses and the systems involved need security measures to protect integrity and availability which will help avoid any malfunctioning to their operations due to the cyber-attacks. There have been several cyber-attack incidents on several businesses such as healthcare, pharmaceutical, water cleaning, and energy sector.

According United States Geological Survey, Aceh is the northwestern part in Indonesia that has been affected by numerous strong earthquakes since 2004 tsunami. These earthquakes have generated massive impact to the buildings around the area, especially for the reinforced concrete based buildings. One of the most important problems to the reinforced concrete is the earthquake-generated crack. In this study, the dataset from the normal and cracked reinforce concrete are collected by taking the normal and cracked images. Several convolutional neural network models are implemented such as LeNet based models. These models are initially applied to recognize either normal or cracked conditions. Eventually, for the last stage, the localization of the crack is visualized by imposing the original images. For the localization, this study also evaluates the relatively smaller and bigger cracks. The results show the higher input image with modified LeNet generates better results compared to the basic model in superimposing the localized crack. The authors would like to express their gratitude to Universitas Syiah Kuala for financial support partly for research through the Professor's research grant No. 03 / UN11.2 / PP / PNBP / SP3 / 2019.

In order to maintain well-rounded competences in today’s global businesses, organisations seek efficient methods to attract the best qualifications and candidates to maintain a high-level, qualified workforce. Currently, organisations in Saudi Arabia receive a tremendous number of job applications which, require manual selection and identification of the best candidates. This is considered to be a time-consuming process that requires personnel and resources. In this paper, we develop a new e-recruitment support system for organisations that can assess the experience, qualifications and suitability of a candidate applying for a specific job in a company or governmental organisation. In the proposed work, a "codeigniter" framework is used which is the main interface platform for the graphical user interface dashboard to filter the input data of all candidates. The proposed system is built experimentally using mySQL and PHP servers. For the quantitative survey, the Cronbach method is implemented to assess the reliability and validity of the survey. The quality achieved is 94%. The system should analyze job requirements and match the skills of potential employees, selecting the best person for short-listing. Hence, the human factor in the selection process can be eliminated.

Before the full release of 5G, 4G will continue to be developed to support many new use cases and applications. By making some modifications and enhancements on its structure, 4G can handle many new features that could be seen as 5G specifics. Such enhancement includes the use of small cells in heterogeneous networks, mobile edge computing, and cloud computing, virtualizations and software-defined networks (SDN), and network slicing. Heterogeneous network supported by computing functionality is one of the most recent added flavor topologies to LTE technology and the upcoming next-generation mobile network. Mobility is yet one of the main challenges in such a system. Allowing the user to surf the world of the data without interruption while in the move between the different cells of the heterogeneous network in the future 5G network is required to keep the high level of the user’s quality of experience. In this paper, new mobility-enhanced schemes are presented for improving the handover performance of a mobile UE with dual connectivity in heterogeneous network.

This paper presents a design and prototype of an IoT-based health and safety monitoring system using MATLAB GUI. This system, which is called the Smart Health and Safety Monitoring System, is aimed at reducing the time, cost and manpower requirements of distributed workplaces. The proposed system is a real-time control and monitoring system that can access on-line the status of consumable devices in the workplace via the internet and prioritise the critically high location that need replenishing. The system dynamically updates the status of all location, such as first aid boxes, earplug dispensers and fire extinguishers. Simulation results of the proposed system gives shorter path, time and cost in comparison to manual maintenance systems.

A human gesture is a non-verbal form of communication and is critical in human-robot interactions. Vision-based gesture recognition methods play a key role to detect hand motion and support such interactions. Hand gesture recognition allows a appropriate, and usable interface between users and devices. Hand gestures can be used for various fields which makes it be able to be implemented for communication and further. Hand gesture recognition is not only useful for people who are hearing-impaired or disabled but also for the people who have experienced a stroke, as they need to communicate with other people using different common essential gestures such as the sign of eating, drink, family and, more. In this paper, a system for recognizing hand gesture based on Convolutional Neural Network (CNN) is proposed. The developed method is evaluated and compared between training and testing modes based on several metrics such as execution time, accuracy, sensitivity, specificity, positive and negative predictive value, likelihood and root mean square. Results show that testing accuracy is 99% using CNN and is an effective technique in extracting distinct features and classifying data.

Improving flexibility and adding complexity are the most driving terms in any computer or telecommunications network and should be carefully considered when proposing new technical solutions for the next (5G) networks. Until the full release of the next generation mobile network; 4G, represented by LTE-A, will continue to develop and improve with new releases launched periodically; each of them demonstrates new added feature(s) that can be seen to continue working with 5G, such enhancing features include small cells in heterogeneous networks, cloud computing, virtualization, software defined networks (SDN), content delivery network (CDN) and cloud radio access networks (CRAN). Resource allocation in cloud computing is one of the major challenges in such systems; this is due to users’ frequent mobility, hence the need for an efficient load balancing method assisted by prediction trait. In this paper we propose and discuss load balancing technique using Hopfield artificial neural network and Radial Basis Function Neural Network for content delivery mechanism in Heterogeneous LTE-A mobile network.

This study evaluates the combination of computational fluid dynamics (CFD) and artificial intelligence (AI) algorithms in heat exchanger application. More specifically, fuzzy logic controller (FLC) is selected in the closed-loop control system. The sensor read by the FLC is based on the reading from the CFD meshing finite element method (FEM) discretization in order to solve the Navier-Stokes (N-S) equation. The controller output is the boundary condition of the CFD system (i.e heat flux). In this study, the CFD result will provide a predicted value approaching the set point. This system utilizes six rules for the controller. In order to evaluate the controller performance, several different set points are set. In addition, the genetic algorithm (GA) is set as the optimizer for the making of the sequence of the Mamdani triangular membership function. Finally, the GA is used to modify the membership function of the FLC. The total evaluation error for the FLC is 0.92 °C of the mean absolute error (MAE). Meanwhile, the GA-based FCL produces 0.78 °C of the MAE.

The emulation of human behavior for autonomous problem solving has been an interdisciplinary field of research. Generally, classical control systems are used for static environments, where external disturbances and changes in internal parameters can be fully modulated before or neglected during operation. However, classical control systems are inadequate at addressing environmental uncertainty. By contrast, autonomous systems, which were first studied in the field of control systems, can be applied in an unknown environment. This paper summarizes the state of the art autonomous systems by first discussing the definition, modeling, and system structure of autonomous systems and then providing a perspective on how autonomous systems can be integrated with advanced resources (e.g., the Internet of Things, big data, Over-the-Air, and federated learning). Finally, what comes after reaching full autonomy is briefly discussed.

One concern to the patients is the off-line detection of pneumonia infection status after using the ventilator in the intensive care unit. Hence, machine learning methods for ventilator-associated pneumonia (VAP) rapid diagnose are proposed. A popular device, Cyranose 320 e-nose, is usually used in research on lung disease, which is a highly integrated system and sensor comprising 32 array using polymer and carbon black materials. In this study, a total of 24 subjects were involved, including 12 subjects who are infected with pneumonia, and the rest are non-infected. Three layers of back propagation artificial neural network and support vector machine (SVM) methods were applied to patients&rsquo, data to predict whether they are infected with VAP with Pseudomonas aeruginosa infection. Furthermore, in order to improve the accuracy and the generalization of the prediction models, the ensemble neural networks (ENN) method was applied. In this study, ENN and SVM prediction models were trained and tested. In order to evaluate the models&rsquo, performance, a fivefold cross-validation method was applied. The results showed that both ENN and SVM models have high recognition rates of VAP with Pseudomonas aeruginosa infection, with 0.9479 &plusmn, 0.0135 and 0.8686 &plusmn, 0.0422 accuracies, 0.9714 &plusmn, 0.0131, 0.9250 &plusmn, 0.0423 sensitivities, and 0.9288 &plusmn, 0.0306, 0.8639 &plusmn, 0.0276 positive predictive values, respectively. The ENN model showed better performance compared to SVM in the recognition of VAP with Pseudomonas aeruginosa infection. The areas under the receiver operating characteristic curve of the two models were 0.9842 &plusmn, 0.0058 and 0.9410 &plusmn, 0.0301, respectively, showing that both models are very stable and accurate classifiers. This study aims to assist the physician in providing a scientific and effective reference for performing early detection in Pseudomonas aeruginosa infection or other diseases.

During surgeries, the amount of used anesthetic depends on the physical conditions of the patient and is immensely critical. The conventionally used BIS Quantro machine which measures the Bispectral Index (BIS) level in order to help doctors administer anesthesia, is quite expensive. In this paper, an economic, accurate and state-of-the-art technique is presented to predict the depth of anesthesia (DoA) via advanced deep learning models using 512 Hz Electrocardiogram (ECG) and 128 Hz Photoplethysmography (PPG). The study is conducted based on signal collected from 50 patients acquired during surgery at National Taiwan University Hospital (NTUH). First, heatmaps of the ECG and PPG signals (individual and combined subplots) are generated using MATLAB by filtering 5 s windows to match the frequency of the BIS Quantro Machine which is 0.2 Hz. Then, various deep learning models comprising 5, 6, 8, 10 and 19 layered CNNs are trained using data of 40 patients and tested using the remaining 10 patients. The heatmap images of ECG and PPG are fed as inputs to the CNN models separately and using two input channels. The best accuracy achieved is 86 % which is attained using 10 layered CNN with Tensorflow backend, with combined ECG and PPG heatmaps as inputs. This study uses inexpensive signals, minimum data reconstruction, minimum memory and timing constrains to achieve a decent accuracy, and so it can be used by even small hospitals.

Microwave energy in combination with AC corona initiated non-thermal plasma is a promising method for removing of NOx and SOx from the exhaust of diesel engine. Recently experimental work shows that a microwave oven (1.8 kW) can be used as an experimental system to research this subject. However, microwave ovens are often based on cyclic mode operation for controlling the power. This type of control is not suitable for the purpose of continuously cleaning the exhaust gas. This study investigates the use of phase control circuit as a regulator for controlling the power in the microwave reactor. The system was tested and its performance was evaluated. It was observed that phase-controlled power regulator could be successfully used for quasi-continuous (fast-switching) power regulation with the maximization of power efficiency. The developed system is used to reduce the emission gases from a marine diesel engine. Different power settings were used to investigate the effect of the power on the reduction efficiency. The same controlled microwave source based non-thermal plasma reactor will be used to investigate the pollutant removal efficiency of the exhaust gas emitted from a 40 kW marine Diesel engine.

In this paper we have deployed the concept of deep learning known as convolutional neural networks (CNN) as we can realize nowadays deep learning is growing in each and every field. Deep learning is executed in each and every platform and its outcome is impressive. On the other hand, the capability and accuracy of deep learning is somehow compared with human beings. We trained CNN to classify either defective or good printed circuit board (PCB). In this experiment we have used 41,387 images, which is divided into 3 different data sets i.e. training, validation and testing. The CNN, which has 60 million parameters and 500,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and two globally connected layers with a final 1000-way softmax. Hence, deep learning via convolution neural networks has been introduced in this paper, which will eventually increase the accuracy and reduce a lot of time and consumption of skilled manpower. According to this preliminary study, we can overall achieve accuracy of above 85% and minimize the count of defective PCB classifying as good. In the near future, we hope that over 95% accuracy can be achieved by using different CNN models like VGGNET, RESNET and GOOGLENET and collecting more PCB image data in order to reduce the consumption of time, manpower and increase the accuracy in quality inspection.

Copyright © 2019 by the authors. Recently, significant developments have been achieved in the field of artificial intelligence, in particular the introduction of deep learning technology that has improved the learning and prediction accuracy to unpresented levels, especially when dealing with big data and high-resolution images. Significant developments have occurred in the area of medical signal processing, measurement techniques, and health monitoring, such as vital biological signs for biomedical systems and noise and vibration of mechanical systems, which are carried out by instruments that generate large data sets. These big data sets, ultimately driven by high population growth, would require Artificial Intelligence techniques to analyse and model. In this Special Issue, papers are presented on the latest signal processing and deep learning techniques used for health monitoring of biomedical and mechanical systems.

The convolutional neural network (CNN) algorithm is one of the efficient techniques to recognize hand gestures. In human&ndash, computer interaction, a human gesture is a non-verbal communication mode, as users communicate with a computer via input devices. In this article, 3D micro hand gesture recognition disparity experiments are proposed using CNN. This study includes twelve 3D micro hand motions recorded for three different subjects. The system is validated by an experiment that is implemented on twenty different subjects of different ages. The results are analysed and evaluated based on execution time, training, testing, sensitivity, specificity, positive and negative predictive value, and likelihood. The CNN training results show an accuracy as high as 100%, which present superior performance in all factors. On the other hand, the validation results average about 99% accuracy. The CNN algorithm has proven to be the most accurate classification tool for micro gesture recognition.

This present study is what we think is one of the first studies to apply Deep Learning to learn depth of anesthesia (DoA) levels based solely on the raw EEG signal from a single channel (electrode) originated from many subjects under full anesthesia. The application of Deep Neural Networks to detect levels of Anesthesia from Electroencephalogram (EEG) is relatively new field and has not been addressed extensively in current researches as done with other fields. The peculiarities of the study emerges from not using any type of pre-processing at all which is usually done to the EEG signal in order to filter it or have it in better shape, but rather accept the signal in its raw nature. This could make the study a peculiar, especially with using new development tool that seldom has been used in deep learning which is the DeepLEarning4J (DL4J), the java programming environment platform made easy and tailored for deep neural network learning purposes. Results up to 97% in detecting two levels of Anesthesia have been reported successfully.

Renewable energy sources are the answer to countless global issues for instance: the increasing risk of climate changes and global warming. Due to the promising benefits of integrating renewable energies based on generation such as: solar and wind sources the future energy markets are encouraged to invest in the available renewable systems. The Kingdom of Saudi Arabia is currently working to fulfil its 2030 vision, which one of the top priorities is investing in renewable energy resources. The western part of Saudi Electrical Company covers number of significant cities, characterised by harsh weather conditions: very high temperatures during the day and high level of humidity during the night time that force the consumers to use air conditioning on daily basis, hence high load, especially in the summer period. This research focuses on integrating renewable energy into transmission network of the western part, with the objective to minimize CO2 emission, through testing the grid load data with capacity of power stations and integrating solar panel and wind turbine platform.

This paper is proposed to reduce power consumption and cost of operation in the Base Band Unit (BBU) pool by switch off number of BBUs that work at low traffic load. Furthermost, this work is introduced new power consumption model for Cloud Radio Access Network (C-RAN) architecture. Two algorithms have been proposed that host in Control Unit (CU) which located in the BBU pool. One the algorithms is to convert the operation processing of switched off BBUs to the remaining ones depending on their valuable capacities. The other one is to control the electronic Smart Switch (SS) in order to exchange the traffic load between the Optical Transmitters (OTXs). The C-RAN is applied using the DWDM/SCM-PON-RoF architecture to decrease the cost effectiveness. The results in system model validation showed a reduction in the BBUs utilization of 84% and 83% in residential and office area respectively. Furthermore, OTXs operation has been reduced by 86% during the 24 hours compared with the conventional RAN network.

With critical importance of medical healthcare, there exist urgent needs for in-depth medical studies that can access and analyze specific physiological signals to provide theoretical support for practical clinical care. As a consequence, obtaining the valuable medical data with minimal cost and impacts on hospital work comes as the first concern of researchers. Anesthesia plays a widely recognized role in surgeries, which attracts people to undertake relevant research. In this paper, a real-time physiological medical signal data acquisition system (PMSDA) for the multi-operating room applications is proposed with high universality of the hospital practical settings and research requirements. By utilizing a wireless communication approach, it provides an easily accessible network platform for collection of physiological medical signals such as photoplethysmogram (PPG), electrocardiograph (ECG) and electroencephalogram (EEG) during the surgery. In addition, the raw data is stored on a server for safe backup and further analysis of depth of anesthesia (DoA). Results show that the PMSDA exhibits robust, high quality performance and efficiently reduces costs compared to previously manual methods and allows seamless integration into hospital environment, independent of its routine work. Overall, it provides a pragmatic and flexible surgery-data acquisition system model with low impact and resource cost applicable to research in critical and practical medical circumstances.